UBS AG
Postfach
8098 Zürich
Tel. +41-44-234 11 11
Corporate Center
Group CFO Management
Tom Naratil
Tel. +41-44-234 24 07
Fax +41-44-234 21 00
tom.naratil@ubs.com
www.ubs.com
Ms. Stephanie Ciboroski
Senior Assisstant Chief Accountant
Division of Corporation Finance
Securities and Exchange Commission
100 F Street, NE
Washington DC 20549-1090
28 February 2013
UBS AG
Form 20-F for the Fiscal Year Ended December 31, 2011
Filed March 15, 2012
Forms 6-K filed July 31, 2012 and October 30, 2012
Response dated September 10, 2012
File No. 001-15060
Dear Ms. Ciboroski:
We respectfully submit the responses set forth below to the Comment Letter dated January 30, 2013 (the “Comment Letter” and individually or collectively “Comments”) from the Staff of the Securities and Exchange Commission relating to (a) the annual report on Form 20-F for the fiscal year ended December 31, 2011 (the “2011 Form 20-F”), (b) the second quarter 2012 report on Form 6-K (the “Q2 2012 Form 6-K”) and (c) the third quarter 2012 report on Form 6-K (the “Q3 2012 Form 6-K”) of UBS AG (“UBS”). For your convenience, we have set out the text of each Comment contained in the Comment Letter, followed by our response. We are in the final stages of preparing our annual report for the fiscal year ended December 31, 2012, which is expected to be authorized for issue by the Board of Directors on March 7, 2013 and then filed on Form 20-F (the “2012 Form 20-F”). Where relevant, we have included in the relevant responses below our current proposed disclosure for the 2012 Form 20-F.
Comment 1
Form 20-F for Fiscal Year Ended December 31, 2011
Composition of credit risk – business divisions, page 120
Wealth Management & Swiss Bank, page 120
We note your response to prior comment 5. Please provide us with the proposed disclosures that you intend to provide in your future filings to address our comment.
We have considered your Comments requesting additional disclosure for our residential real estate loan portfolio with regard to our process for loan-to-value (“LTV”) ratios at the origination of such loans, quantitative information on our current LTV ratios and the process for calculating current LTV ratios. Our current proposed disclosure on this matter, in the Composition of credit risk – business divisions section of the 2012 Form 20-F, is below. We believe that this enhanced disclosure addresses these Comments.
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 2 of 22 |
“Our largest loan portfolio is our mortgage loan portfolio, which principally comprises loans within Switzerland which are secured by residential and commercial real estate. These mortgage loans mainly originate from our Retail and Corporate business but also include mortgage loans originating from our Wealth Management business. The majority of these mortgage loans relate to residential properties that the borrower either occupies or rents out and are full recourse to the borrower.
We use a scoring model as part of a standardized front-to-back process to support credit decisions for the origination or modification of all Swiss mortgage loans. The two key factors within this model are an affordability calculation relative to gross income and the loan-to-value ratio (LTV). The calculation of affordability takes into account interest payments, minimum amortization requirements, potential property maintenance costs and, in the case of properties expected to be rented out, the level of rental income. Interest payments are estimated using a predefined framework, which takes into account the potential for significant increases in interest rates during the lifetime of the loan.
Approximately 70% of the Swiss residential mortgage loan portfolio relates to properties occupied by the borrower. For such mortgage loans, the maximum LTV allowed within the standard approval process is 80%. This is reduced to 60% in the case of vacation properties and luxury real estate. The value assigned by UBS to each property is based on the lowest value determined based on an internal valuation, the purchase price and, in some cases, an additional external valuation. Valuations of owner-occupied real estate are reviewed and updated throughout the lifetime of the loan, typically using real estate price indices. The average LTV ratio of this portfolio was approximately 55% at 31 December 2012, compared with 58% at 31 December 2011. Over 99% of the aggregate amount of mortgage loans within this portfolio would continue to be covered by the real estate collateral even if the value assigned to that collateral were to decrease by 20%. Furthermore, these loans are full recourse to the borrower. The average LTV for 2012 of newly originated loans in this portfolio was 63%.
Approximately 30% of the Swiss residential mortgage loan portfolio relates to properties rented out by the borrower. For such mortgage loans, the maximum LTV allowed within the standard approval process ranges from 60% to 80%, depending on the type of property, the age of the property and the amount of any renovation work required. LTVs are reviewed and updated periodically throughout the lifetime of the loan. The rental income from properties is reviewed, at a minimum, once every three years, but indications of significant changes in the amount of rental income or the level of vacancy rate can trigger an interim reappraisal. The average LTV ratio of this portfolio was approximately 58% at 31 December 2012, compared with 59% at 31 December 2011. Over 99% of the aggregate amount of mortgage loans within this portfolio would continue to be covered by the real estate collateral even if the value assigned to that collateral were to decrease by 20%. Furthermore, these loans are full recourse to the borrower. The average LTV for 2012 of newly originated loans in this portfolio was 56%.”
Comment 2
Consolidated Financial Statements, page 284
Note 1 – Summary of significant accounting policies, page 297
a) Significant accounting policies, page 297
10) Loans and Receivables – Renegotiated Loans, page 301
We note your response to prior comments 14 and 15. Please address the following:
• | Revise your disclosure in future filings to clarify the differences between loans renegotiated on concessionary terms and loans renegotiated on a non-concessionary basis in terms of nonperforming classification and method of impairment measurement. |
• | Tell us and revise future filings to disclose whether you monitor or segregate these loans from other non-renegotiated loans within the general collective loan assessment for purposes of measuring credit losses. If you do not segregate these loans from loans that were not previously renegotiated, please tell us whether you otherwise take into consideration the fact that these loans have been renegotiated when measuring credit losses. |
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 3 of 22 |
We have considered your Comments with regard to clarifying the differences between loans renegotiated on concessionary terms and loans renegotiated on a non-concessionary basis—in terms of non-performing classification and method of impairment measurement. Additionally, we have considered how we treat and monitor these loans within the general collective loan assessment for purposes of measuring credit losses. Consequently, our current proposed disclosure on this matter in Note 1, the Summary of significant accounting policies section of the 2012 Form 20-F, is as follows:
“A renegotiated loan is defined as a loan that has been subject to restructuring, or for which additional collateral has been requested that was not contemplated in the original contract.
Typical key features of terms and conditions granted through renegotiation to avoid default include the provision of special interest rates, postponement of interest or amortization payments, modification of the schedule of repayments or amendment of loan maturity. There is no change in the effective interest rate following a renegotiation.
If a loan is renegotiated with concessionary conditions (i.e. new terms and conditions are agreed which do not meet the normal market criteria for the quality of the obligor and the type of loan), the position is still classified as non-performing and is rated as being in counterparty default. It will remain so until the loan is collected or written off and will be assessed for impairment on an individual basis.
If a loan is renegotiated on a non-concessionary basis (e.g. additional collateral is provided by the client, or new terms and conditions are agreed which meet the normal market criteria for the quality of the obligor and the type of loan), the loan will be re-rated using our regular rating scale. In these circumstances, the loan is removed from impaired status and therefore included in our collective assessment of loan loss allowances. For the purposes of measuring credit losses, these loans are not segregated within the collective loan loss assessment from other loans which have not been renegotiated. Management regularly reviews all loans to ensure that each of the criteria contained in the loan agreement continues to be met and that future payments are likely to occur.
In those situations where the restructuring of a loan results in a fundamental change in the terms and conditions of the loan, the original loan is derecognized and a new loan is recognized. A change is considered fundamental if the present value of the contractual cash flows (as a proportion of notional) has been changed by 10% or more, or there has been a significant change in the risk profile of the instrument. If a loan is derecognized in these circumstances, the new loan is measured at fair value at initial recognition. Any allowance taken to date against the original loan is eliminated and is not attributed to the new loan. Consequently, the new loan is not considered impaired and is included within the general collective loan loss assessment for the purpose of measuring credit losses.”
Comment 3
Note 16 – Goodwill and intangible assets, page 336
We note your response to prior comment 16, including the statement that you do not specifically identify the assets and liabilities that make up the excess capital. Please confirm that the carrying value of each cash generating unit (CGU) includes only those assets that can be attributed directly, or allocated on a reasonable and consistent basis, to each CGU and provide a robust explanation of your asset allocation process. In this regard, clarify whether you are able to specifically identify the assets that remain in the Corporate Center after allocations (or conversely, whether you are able to specifically identify which assets have been allocated out to the CGUs). If not, please tell us how you concluded that you met the criteria in paragraphs 76 of IAS 36.
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 4 of 22 |
As noted in the Capital Management section of our 2011 Form 20-F, equity is attributed to the businesses and the Corporate Center (“CC”) after considering their risk exposure, risk-weighted assets (“RWA”) usage, asset size, goodwill and intangible assets. Thus in order to attribute equity under the equity attribution framework (“EAF”), the assets of each segment (or cash-generating unit (“CGU”)) and the CC are identified. Because we identify the assets attributable to each CGU and the CC, we can confirm that the carrying amount of each respective CGU and the CC includes only those assets that can be attributed directly or allocated on a reasonable and consistent basis to it.
We attribute assets directly to the CGUs and the CC based on the way assets are managed internally, meaning that assets are attributed to the business division (or the CC) that is responsible for managing them and their performance. Hence, that approach is aligned with our segment reporting pursuant to IFRS 8 Operating Segments as most recently disclosed in our fourth quarter 2012 report on Form 6-K (“the Q4 2012 Form 6-K”).
As we can identify the assets in the CC, for the purpose of our goodwill impairment testing, we evaluate them to determine whether they represent corporate assets, as described in IAS 36.100-101. Corporate assets in the CC are identified and allocated to the CGUs on a reasonable and consistent basis. Allocated corporate assets consist of property and equipment, which have been allocated based on actual rates of usage incurred by each CGU. Other assets in the CC are not deemed to represent corporate assets because they generate cash flows independently of the other CGUs and thus are not allocated (e.g. our multi-currency portfolio of unencumbered, high quality, short-term financial assets that are managed by Group Treasury). The Legacy Portfolio financial assets in CC also are not allocated for that same reason.1
As a result of the process described above, we are able to specifically identify the assets that remain in the CC after allocations. We believe that our attribution and allocation methodology results in an appropriate basis for determining the carrying amount of each CGU in accordance with IAS 36.76–79.
Comment 4
Furthermore, to the extent that you do not identify the specific assets and liabilities directly attributed to each CGU, please tell us how you are able to estimate the cash flows from such CGUs. In your response, please address why you believe your methodology complies with the guidance in paragraph 77 of IAS 36.
As explained in our response to Comment 3, we are able to identify the specific assets for each CGU. Thus we have the requisite information to determine that our attribution and allocation methodology results in an appropriate basis for determining both the recoverable amount and the carrying amount of our CGUs. Hence, we believe that our methodology complies with the guidance in IAS 36.77.
Comment 5
In your response dated September 16, 2011 you noted that one potential shortfall of the equity attribution framework was that the attribution process will not always result in aggregated attributed equity at the segment level being the same as equity attributed to your shareholders. You further recognized that adjustments would need to be made to ensure that the carrying amount for purposes of the goodwill impairment test was not less than the equity attributable to shareholders and therefore the net carrying amount of your assets and liabilities, which you believed was consistent with the premise and requirements of paragraphs 76-79 of IAS 36. Please tell us how your current equity attribution framework, including how it is used in determining the carrying value of your CGUs for purposes of goodwill impairment testing is consistent with your prior response.
1 | The Legacy Portfolio was transferred to CC before the end of 2011; its assets consist of financial instruments. |
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 5 of 22 |
Please also refer to our response to Comment 3, where we outline our attribution and allocation methodology which results in an appropriate basis for determining the carrying amount of each CGU in accordance with IAS 36.76–79.
The following table presents the difference between the average aggregated attributed equity at the segment level and the average equity attributable to shareholders for the past five quarters.
CHF Billions | Q4 12 | Q3 12 | Q2 12 | Q1 12 | Q4 11 | |||||||||||||||
Average attributed equity at the segment level | 45.5 | 50.8 | 51.3 | 55.5 | 56.5 | |||||||||||||||
Difference | 1.5 | 2.8 | 2.7 | (2.2 | ) | (3.9 | ) | |||||||||||||
|
|
|
|
|
|
|
|
|
| |||||||||||
Average equity attributable to shareholders | 47.0 | 53.6 | 54.0 | 53.3 | 52.6 | |||||||||||||||
|
|
|
|
|
|
|
|
|
|
Source: 2012 Quarterly Reports (Quarters prior to Q4 2012 have not been restated for the adoption of IAS 19 RevisedEmployee Benefits).
We note that the allocation of those differences to the CGUs on a reasonable and consistent basis (for example, on the basis of relative total assets or relative recoverable amounts) would not have changed the results of our goodwill impairment testing as of 31 December 2011. In addition, it would not have changed the timing of the goodwill impairment recognized for the Investment Bank in Q3 2012.
We will continue to monitor whether the difference between total attributed equity and IFRS equity attributable to shareholders could potentially impact the results of our impairment tests.
Comment 6
Note 21 – Provisions and contingent liabilities, page 341
We note your response to the first bullet of prior comment 17. You state that “the classes of provisions are the same as the classes of contingent liabilities; however, contingent liabilities with respect to operational risks, restructuring, loan commitments and guarantee, and other are either not present or not deemed to merit further disclosure in the notes to the financial statements.” It appears you are saying that for the above classes, the contingent liabilities are either immaterial or do not exist. If true, affirmatively state that in your footnote disclosure in future filings for each class, as this cannot be inferred from your current disclosure.
Our current proposed disclosure in the footnotes to the Provisions table in the 2012 Form 20-F is as follows:
“Further information on provisions and contingent liabilities that are more than remote in respect of Litigation, regulatory and similar matters, as a class, is included in section (b) of this Note. Further information on the nominal principal amount of Loan commitments and guarantees, representing our maximum exposure to loss, is disclosed in Note 29Measurement categories of financial assets and financial liabilities. There are no material contingent liabilities associated with the other classes of provisions.”
Comment 7
We note your response to the second bullet of prior comment 17. You appear to state that you have not provided the disclosure required by paragraph 85(a) regarding the timing of any outflows of economic benefits and the disclosures required by paragraph 85(b) for your restructuring provision based on your determination that the information was not material. We note that your restructuring provision increased by CHF 393 million in 2011, which represents over 7% of your pre-tax income. Please tell us why you believe the above disclosures are immaterial in the context of your restructuring provision. Alternatively, please tell us how you will revise future filings to provide the above disclosures.
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 6 of 22 |
As disclosed in Note 1 27) Provisions on page 309 of our 2011 Form 20-F, restructuring provisions are recognized when a detailed and formal restructuring plan has been approved and a valid expectation has been raised that the restructuring will be carried out, either through commencement of the plan or announcements to affected employees. Such provisions tend to be utilized within a relatively short period of time after the balance sheet date and the uncertainty with regard to the amount of outflows is significantly lower than in the case of, say, litigation provisions. For example, severance provisions triggered by restructuring plans are generally calculated based on a formulaic approach that takes into account the set terms of the redundancy packages and the number of employees affected.
Given the short term nature of the resulting outflows and the limited uncertainty of the amounts recognized, we have taken the position that additional information would not “influence the economic decisions that users make on the basis of the financial statements” [IAS 1.7]. Nevertheless, following the accelerated implementation of our strategy announced in October 2012 and as disclosed in our Q4 2012 Form 6-K, we expect to incur significant restructuring charges over the coming quarters as we continue to execute on our strategy and we implement future cost reduction initiatives.
In consideration of the above, we are proposing the following disclosure to be included within the Provisions and contingent liabilities Note of our 2012 Form 20-F:
“UBS’s restructuring provisions primarily relate to onerous lease contracts and severance amounts. The utilization of onerous lease provisions is driven by the maturities of the underlying lease contracts, which cover a period of up to xx years. Severance related provisions are utilized within a short time period, usually within six months, but potential changes in amount may be triggered when natural staff attrition reduces the number of people affected by a restructuring and therefore the estimated costs.”
Comment 8
We note your response to the second bullet of prior comment 17. You state that you have provided the disclosure required by paragraph 85(a) regarding the expected timing of any outflows of economic benefits for your litigation and regulatory matters provision class in Note 21 b). We are unable to locate this disclosure. Please direct us to the sentences in that footnote that you believe satisfy this disclosure requirement.
The following sentence on page 342 of the 2011 Form 20-F reflects our expectations with regard to the timing of outflows and the related uncertainties:“Such cases are subject to many uncertainties, and the outcome is often difficult to predict, including the impact on operations or on the financial statements, particularly in the earlier stages of a case.”That statement from Note 21 b) highlights the significant uncertainties relating to provisions and contingent liabilities for Litigation, regulatory and similar matters, which limit the amount of information on which we can base expectations with regard to the timing of outflows. We recognize the potential to enhance the clarity of the disclosure provided in relation to paragraph 85(a) and propose amended wording as per our response to Comment 9 below.
Please refer to our response to Comment 10 for a description of factors affecting the expected timing of the outflow of economic benefits for our litigation, regulatory and similar matters provision class.
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 7 of 22 |
Comment 9
We note your response to the second bullet of prior comment 17. You state that you have provided the disclosure required by paragraph 85(b) about the uncertainties surrounding the amount or timing of the outflow of economic benefits for your litigation and regulatory matters provision class in Note 21 b). We note your disclosure that “[s]uch cases are subject to many uncertainties…” However, other than the uncertainty regarding whether you will enter into a settlement, as described in the following sentence in your disclosures, we are unable to locate any further description of the “many uncertainties” to which you refer. We also note your disclosure in the following paragraph regarding the speculative nature of your legal assessments due to unique fact patterns, novel legal theories, early stages of adjudications, and unquantified damages; however, this disclosure appears to be in the context of explaining why it is impracticable to estimate the financial effect of this class’s contingent liability. However, the practicability accommodation is not available for the disclosures required by paragraph 85(b). Accordingly, please direct us to the sentences in that footnote that you believe satisfy this disclosure requirement.
As noted in our response to Comment 8 above, the uncertainties indicated in the second paragraph of Note 21b) of the 2011 Form 20-F relate to both the provisions recognized as well as to this class of contingent liabilities, given the similarity of the triggering circumstances. We have recognized that it was not clear in our disclosure that this information related both to provisions and contingent liabilities and, in the Q4 2012 Form 6-K, we amended this disclosure as follows:
“Such matters are subject to many uncertainties and the outcome is often difficult to predict, particularly in the earlier stages of a case.The uncertainties inherent in all such matters affect the amount and timing of any potential outflows for both matters with respect to which provisions have been established and other contingent liabilities have been identified.”
Please refer to our response to Comment 10 for further discussion regarding the uncertainties surrounding the amount or timing of the outflow of economic benefits for our litigation, regulatory and similar matters provision class.
Comment 10
We note your response to prior comments 18 and 19. You state that “paragraph 87 of IAS 37 places a heavy emphasis on the nature of the items in determining which of them can be aggregated to form a class.” You also state that you believe the nature of the provision is the “cause giving rise to the provision” and that “when such causes are deemed similar because general statements about their nature and other characteristics of the items described in [paragraphs 85 and 86 of IAS 37] provide useful information, they are aggregated … to form class.”
However, paragraph 87 of IAS 37 does not base aggregation on an entity’s ability to merely make similar “general [emphasis added] statements” about the disclosure requirements in paragraphs 85 and 86 of IAS 37. It states that “it is necessary to consider whether the nature of items is sufficiently similar for asingle [emphasis added] statement about them to fulfill the requirements of paragraphs 85(a) and (b) and 86(a) and (b)” of IAS 37.
Accordingly, as previously requested in prior comment 18, please provide us with a detailed analysis of how you determined that the individual items in the litigation and regulatory matters class may be aggregated to form a single class. For each of the 14 matters discussed in Note 21 b), please tell us the following:
• | The expected timing of cash out flows and how the expected timing for each item is sufficiently similar that you are able to make a single statement for all matters in the class. |
• | The uncertainties in the amount and timing of the outflow of an economic benefit. Explain whether or not the uncertainties will increase or decrease the timing or the amount of the outflow. In addition, please discuss how the uncertainties in the amount and timing for each item are sufficiently similar that you are able to make a single statement for all matters in the class. |
• | The estimated financial effects of the contingent liabilities and how the financial effects for each item are sufficiently similar that you are able to make a single statement for all matters in the class. In this regard, we note in your response to prior comment 21 and 22 that it is not practicable to estimate the financial effect of the contingent liability for some matters, but that you are able to estimate it for other matters. We also note from your disclosures that you are unable to estimate the possible financial effect due to some combination of the reasons you provide – in other words, that there is no consistent reason that applies to all matters in this class. We further note in your response to prior comment 21 that you have recognized a provision for some of the matters in this class, but not for others. These differences suggest that the uncertainties associated with the individual matters in this class may not be sufficiently similar for a single statement to be made about them. |
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 8 of 22 |
We have provided the following detailed analysis regarding our determination that the individual items considered as Litigation, regulatory and similar matters are appropriately aggregated to form a single class pursuant to IAS 37.86.
As background, as we indicated in our prior response, the matters included in the Litigation, regulatory and similar matters class have a similar cause. UBS operates in a legal and regulatory environment that exposes it to significant litigation and similar risks arising from disputes and regulatory proceedings. Those matters are unified both by their sources and by their nature. Possible liabilities in this class arise from assertions or threatened assertions by governmental authorities or private claimants that we failed to comply with or otherwise have liability under a law or regulation, or that we failed to satisfy a contractual or other duty to another party or parties, including duties to disclose information or risks, to ensure the suitability of product offerings or to provide financing or services. As a class, they are subject to characteristic uncertainties and factors making outcomes difficult to predict in most circumstances. They are ultimately subject to decisions by courts, regulatory authorities, arbitrators, administrative agencies or bodies or other decision makers not within our control. In attempting to assess potential actions of such decision makers and their consequences, or to assess potential negotiated settlements, we evaluate, among other things and if relevant, the nature of the claim, the underlying facts, our procedural posture and litigation history, our experience and the experiences of others in similar matters, prior or ongoing settlement discussions with the claimant, prior actions of the decision making body or person, available indemnities and the opinions and views of legal counsel. Matters in this class are subject to some or all of the following types of identified uncertainties: (a) settlement uncertainty: we may decide to settle in order to avoid the expense, management distraction or reputational implications of continuing to contest a matter (even if we believe that we should be exonerated); (b) legal/regulatory uncertainty: the treatment of a claim under the law or regulation may require speculative assessments or proceedings that involve novel legal theories, making it difficult to evaluate the probability of future rulings of courts, other tribunals or regulators; (c) discovery/information uncertainty: claims are often asserted before facts and evidence pertinent to a matter are known or considered by claimants or by us, and damages (if any) may not have been quantified by claimants, may not be known to claimants or to us, or may be based on inaccuracies or incomplete information, particularly at the early stages of a matter; and (d) estimation uncertainty: unlike other challenging measurement issues, the use of statistical or other quantitative analytical models is of less relevance for these matters than the application of judgment to the inherent uncertainties referred to above. We also note that matters in this class may be subject to unknown uncertainties.
These common characteristics unify these matters into a class and support our conclusion that our aggregation of matters as a single class under Litigation, regulatory and similar matters is appropriate under IAS 37.87. Information on individual matters within a class is not required by paragraphs 85 and 86 of IAS 37. We nevertheless do provide certain known information on matters within the class because we believe this is useful to investors, but do not provide (a) information for which the disclosure is expected to prejudice seriously our position with other parties in the matter or (b) information that is subject to confidentiality obligations that preclude such disclosure.
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 9 of 22 |
We continue to strive to enhance our disclosures and, in order to enhance our disclosure of information on this class of provisions, we propose including additional information in the Provisions table in the 2012 Form 20-F to show the provisions amount for this class disaggregated by business division, including the Corporate Center. We believe this additional information will facilitate the user’s understanding of this class of provisions and assist in aligning this information to relevant disclosure in the Financial and operating performance section of our 2012 Form 20-F.
An assessment of the application of the specific IAS 37 disclosure requirements to this class follows:
IAS 37.85(a) A brief description of the nature of the obligation and the expected timing of any resulting outflows of economic benefits for provisions
We consider that the brief description of Litigation and regulatory matters provided in Note 21 of the 2011 Form 20-F is sufficient in describing the overall nature of those matters and therefore this class. In addition, as noted in Comment 9 above, we have further enhanced our disclosures in this area, most recently in our Q4 2012 Form 6-K. With regard to the expected timing of any resulting outflows, we note that such timing depends on some or all of the uncertainties previously described. As a consequence and where we believe that a reliable estimate of the obligation can be made, our provision generally reflects a heavier emphasis on the overall amount of the expected outflow and less emphasis on the expected timing of the expected outflow. Litigation, regulatory and similar matters as a class does not lend itself to standard distribution curves in estimating the timing of outflows due to the uncertainties that we have described. These factors establish the basis to make a single statement regarding the timing of any resulting outflows. Further, we note that when settlements or other resolutions are determined, such that no uncertainty remains with respect to the timing or amount of resulting outflows, the amounts are reclassified toOther liabilities.
IAS 37.85(b) An indication of the uncertainties about the amount or timing of those outflows for provisions
As described above, we believe that litigation, regulatory and similar matters as a class is subject to some or all of the following types of uncertainties: (a) settlement uncertainty, (b) legal/regulatory uncertainty, (c) discovery/information uncertainty and (d) estimation uncertainty. All matters are subject to settlement uncertainty and settlements generally result in expected outflows occurring sooner than they would for matters that are only resolved through a full and complete judicial, regulatory, or other process. All matters are subject to legal/regulatory uncertainty; those uncertainties make timing less certain because it is often unclear how long it will take to resolve the uncertainty. As a generalization, one can say that legal and regulatory uncertainty tends to delay the timing of expected outflows, but unexpected legal or regulatory decisions may accelerate the timing of expected outflows. All matters are subject to discovery/information uncertainty, which depending on the nature of the information discovered or clarified may either decrease or increase the timing of expected outflows (e.g. result in a decision to settle or litigate a particular matter). All matters are subject to estimation uncertainty, which has an effect similar to discovery/information uncertainty, meaning that more precise estimates may either decrease or increase the timing of expected outflows (e.g. result in a decision to settle or litigate a particular matter). As such uncertainties are common to all matters, we believe that they provide an appropriate basis on which to make a single statement regarding the amount or timing of those outflows.
IAS 37.86(a) and (b) An estimate of its financial effect and an indication of the uncertainties relating to the amount and timing of any outflow for contingent liabilities
Contingent liabilities related to litigation, regulatory and similar matters are subject to the same set of uncertainties as explained in the analysis under IAS 37.85(a) and 85(b) above. We believe that those characteristics establish the basis to make a single statement regarding contingent liabilities for this class and in our Q4 2012 Form 6-K we enhanced our disclosure to include the following statement:
“It is not practicable to provide an aggregate estimate of liability for our litigation, regulatory and similar matters as a class of contingent liabilities. Doing so would require us to provide speculative legal assessments as to claims and proceedings that involve unique fact patterns or novel legal theories, which have not yet been initiated or are at early stages of adjudication, or as to which alleged damages have not been quantified by the claimants.”
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 10 of 22 |
Comment 11
We have considered your response to prior comment 19. You state that disaggregation of loan repurchase demands into a separate class would not “meaningfully augment the information presented” because you disclose the amount of the provision. However, we note that disaggregation would result in additional disclosures, including a rollforward of the changes in the provision. Based on the information disclosed in your filing it continues to be unclear why you have aggregated loan repurchase demands within your litigation and regulatory matters provision class. Please provide us with a comprehensive analysis regarding the similarity of the nature, expected timing, uncertainties in amount and timing, and estimated financial effect, as described above, to support your conclusion.
The detailed analysis incorporated within our response to Comment 10 above applies to loan repurchase demands to the same extent as it does to the other matters within this class. We note in particular that most of the loan repurchase demands that we have received are now the subject of litigation, emphasizing further that loan repurchase demands should be considered in a consistent manner to our other matters subject to litigation or potential litigation, and not disaggregated as a separate class. You mention in Comment 11 that, if loan repurchase demands were disaggregated, additional disclosures would include a roll forward of the changes in the related provision. In order to enhance information for users, we propose including a roll forward table in the 2012 Form 20-F showing the movements in the period for the provision we have established with respect to the matters encompassed within the item “Claims related to sales of residential mortgage-backed securities and mortgages”.
Comment 12
We note your responses to prior comments 17, 21, 22 and 23 with respect to your application of paragraphs 91 and 92 of IAS 37. We will further consider this information with the information currently requested.
We appreciate the additional consideration of our prior responses.
Comment 13
Note 23 – Derivative instruments and hedge accounting, page 351
Derivatives transacted for trading purposes – Detailed example: Credit derivatives, page 354
We note your response to prior comment 25. Please provide us with the proposed disclosures that you intend to provide in your future filings to address our comment.
We have considered your suggestions for enhanced disclosure of our credit derivative exposures and on this basis have determined that the most useful enhancements would be to include further information by counterparty, by type of derivative and by primary purpose. Our current proposed disclosure on this matter in the 2012 Form 20-F is presented below:
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 11 of 22 |
Note 25x Credit derivatives by counterparty
Protection bought | Protection sold | |||||||||||||||||||||||
CHF billion | Fair value: PRV | Fair value: NRV | Notional values | Fair value: PRV | Fair value: NRV | Notional values | ||||||||||||||||||
Broker-dealers | X | X | X | X | X | X | ||||||||||||||||||
Banks | X | X | X | X | X | X | ||||||||||||||||||
Central clearing counterparties | X | X | X | X | X | X | ||||||||||||||||||
Other | X | X | X | X | X | X | ||||||||||||||||||
|
|
|
|
|
|
|
|
|
|
|
| |||||||||||||
Total 31 December 2012 | X | X | X | X | X | X | ||||||||||||||||||
|
|
|
|
|
|
|
|
|
|
|
| |||||||||||||
Total 31 December 2011 | X | X | X | X | X | X |
Note 25x Credit derivatives - by type of instrument
Protection bought | Protection sold | |||||||||||||||||||||||
CHF billion | Fair value: PRV | Fair value: NRV | Notional values | Fair value: PRV | Fair value: NRV | Notional values | ||||||||||||||||||
Single name credit default swaps | X | X | X | X | X | X | ||||||||||||||||||
Multi-name index linked credit default swaps | X | X | X | X | X | X | ||||||||||||||||||
Multi-name other credit default swaps | X | X | X | X | X | X | ||||||||||||||||||
Total return swaps | X | X | X | X | X | X | ||||||||||||||||||
Credit default swap options | X | X | X | X | X | X | ||||||||||||||||||
|
|
|
|
|
|
|
|
|
|
|
| |||||||||||||
Total 31 December 2012 | X | X | X | X | X | X | ||||||||||||||||||
of which: credit derivatives related to economic hedges | X | X | X | X | X | X | ||||||||||||||||||
of which: credit derivatives related to market making | X | X | X | X | X | X | ||||||||||||||||||
|
|
|
|
|
|
|
|
|
|
|
| |||||||||||||
Total 31 December 2011 | X | X | X | X | X | X |
Comment 14
Additional disclosure required under SEC regulations, page 439
D – Information required by industry guide 3, page 445
Impaired and non-performing loans, page 454
We note your response to prior comment 26. Please provide us with the proposed disclosures that you intend to provide in your future filings to address our comment.
In response to your comment, we have considered the disclosure regarding impaired and non-performing loans in our Annual Report, including the significant disclosure that is included on impaired loans from an IFRS perspective in the “Risk, treasury and capital management” section. As non-performing loans are a sub-set of our overall IFRS impaired loan population, we will remove the “Other impaired loans” and add a reference to the “Risk, treasury and capital management” section of our Annual Report, where detailed information about the Group’s entire portfolio of impaired loans (on an IFRS basis) is disclosed. We will further provide a reference to Note 1 of our financial statements, in which we disclose detailed information about the factors we consider indicative of impairment, in addition to non-performance. By linking together the SEC guide 3 disclosure in our annual reports to the extensive disclosure we provide on credit risk management elsewhere in both the annual and quarterly reports, we believe this will prevent confusion and will enhance our disclosure overall. The following revised SEC Industry guide 3 disclosure is proposed for the 2012 Form 20-F:
“A loan (included in Due from banks or Loans) is classified as non-performing: 1) when the payment of interest, principal or fees is overdue by more than 90 days and there is no firm evidence that it will be made good by later payments or the liquidation of collateral; 2) when insolvency proceedings have commenced; or 3) when obligations have been restructured on concessionary terms.
For IFRS reporting purposes, UBS’s definition of impaired loans is more comprehensive, covering both non-performing loans and other situations where objective evidence indicates that the Group may be unable to collect all amounts due. Refer to “Impairment and default – distressed claims” in the “Risk, treasury and capital management” section for comprehensive information about the Group’s impaired loans, of which non-performing loans are a component. Also, see “Note 1” in the “Financial Information” section of this report for a discussion of the various risk factors that UBS considers to be indicative of impairment.
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 12 of 22 |
The table below provides an analysis of the Group’s non-performing loans:
CHF million | 31.12.12 | 31.12.11 | 31.12.10 | 31.12.09 | 31.12.08 | |||||||||||||||
Non-performing loans: | ||||||||||||||||||||
Domestic | 1,121 | 1,199 | 1,164 | 1,462 | 1,431 | |||||||||||||||
Foreign | 395 | 329 | 563 | 3,940 | 3,272 | |||||||||||||||
|
|
|
|
|
|
|
|
|
| |||||||||||
Total non-performing loans | 1,516 | 1,529 | 1,727 | 5,402 | 4,703 | |||||||||||||||
|
|
|
|
|
|
|
|
|
| |||||||||||
CHF million | 31.12.12 | 31.12.11 | 31.12.10 | 31.12.09 | 31.12.08 | |||||||||||||||
Gross interest income that would havebeen recorded on non-performingloans: | ||||||||||||||||||||
Domestic | 8 | 10 | 11 | 13 | 16 | |||||||||||||||
Foreign | 3 | 9 | 35 | 89 | 7 | |||||||||||||||
|
|
|
|
|
|
|
|
|
| |||||||||||
Interest income included in Net profitfor non-performing loans: | ||||||||||||||||||||
Domestic | 28 | 29 | 35 | 41 | 32 | |||||||||||||||
Foreign | 6 | 6 | 19 | 30 | 6 |
UBS does not, as a matter of policy, typically restructure loans to accrue interest at rates different from the original contractual terms or reduce the principal amount of loans. Refer to the “Credit risk” section of this report for more information. Instead, specific loan allowances are established as necessary. Unrecognized interest related to restructured loans was not material to the results of operations in 2012, 2011, 2010, 2009 or 2008.”
Comment 15
Form 6-K filed July 31, 2012
UBS business divisions and corporate center
Corporate Center – Core Function, page 43
Results: 2Q vs. 1Q, page 44
We note your response to prior comment 27. Please provide us with a more detailed explanation of the specific refinement that was made to your methodology to determine the change in fair value when the interest rate swaps do not have a fair value of zero at inception.
Our macro cash flow hedge (MCFH) models are structured according to the IAS 39 Implementation Guidance (IG) F6.1 – F6.3 as part of our long established banking book risk management framework for interest rate risk. The specific refinement that was made in Q2 2012 was a refinement of the hedge effectiveness calculation. Under the original calculation, the retrospective hedge effectiveness test compared changes in the cash flows of the hedged forecasted transactions with changes in the cash flows of the floating rate legs of the interest rate swaps designated in the model. This calculation assumed that the ineffectiveness from the fixed rate leg of the interest rate swaps was zero, but such an assumption only holds when the interest rate swaps have a fair value of zero at designation (i.e. there is no positive or negative replacement value at designation). The actual fair value at designation (known as the take on replacement value, or “ToRV”) had been monitored separately and was never sufficiently significant to have a material effect on the cumulative life to date hedge effectiveness test or to change our conclusion that the hedges had been highly effective in all periods. This refinement was made to incorporate the ToRV effect within the retrospective hedge effectiveness testing. The refined retrospective hedge effectiveness calculations incorporate within the model the present value change of the remaining excess (or deficit) of the fixed leg compared to the floating leg cash flows of an interest rate swap at the hedge designation date, resulting in a more precise measure of hedge effectiveness.
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 13 of 22 |
Comment 15, continued
Specifically, tell us whether this was due to a change in how you determine the fair value of the hedging instrument, at inception or subsequent, and clarify your old and new methodology so we can better understand the change.
Additionally, please clarify whether there was any change in fair value methodology for any other derivative instruments, in hedging relationships or not in hedging relationships, which did not have a fair value of zero at inception, based on your determination to change the methodology to determine the change in fair value when interest rate swaps do not have a fair value of zero at inception.
This refinement was not due to a change in how we determined the fair value of interest rate swaps. The fair value methodology for other derivative instruments, whether in hedge accounting relationships or otherwise, was not affected by this refinement. There was no change made to our method of effectiveness testing and for information on the old and the new calculation, please see the response above.
Comment 15, continued
As part of your response, provide us with background information on the hedge itself (including how many hedges were affected), the method by which hedge ineffectiveness is assessed and measured, and the driver for the change in methodology.
The MCFH models contain interest rate swaps that are entered into for asset liability management purposes. These swaps are designated in macro hedge accounting relationships as hedging instruments for interest rate risk inherent in the hedged cash flows. Further information regarding the swaps within our MCFH models can be found in Note 23 of our 2011 Form 20-F. Five of the MCFH models were affected by the refinement of the calculation described above. Please see response above for further information on the hedge effectiveness calculation, together with the refinement to that calculation.
Comment 15, continued
Provide sufficient details around the change that was made to the valuation of the hedging instrument, the reason(s) that such changes were made, and clarification on why only certain hedging instruments were impacted.
As explained above, there were no changes made to the actual valuation of the hedging instrument. This refinement only had a financial effect for certain models because of the relative financial position of the hedge accounting relationship. For three models, the cumulative life-to-date change in fair value of the floating legs of the designated interest rate swaps, including the effect of the ToRV, continued to be less than the change in fair value of the variable hedged cash flows. This financial position results in an under-hedge and the new calculation only affected the hedge ratio (i.e. the 80–125% threshold) and did not result in any income statement effect. Conversely, for two models the direct incorporation of the ToRV resulted in an over-hedge, and the effect of this over-hedge was booked to the income statement on a life to date basis.
Comment 16
As a related matter, despite your assertion that the prior period hedge ineffectiveness error was not material to any current or prior period, the actual nature of this error remains unclear. Accordingly, please provide us with a detailed response that clearly describes the nature of the error, the periods that were impacted, and how and when it was discovered. Please also tell us how you considered whether this error impacted your assessment of hedge effectiveness considering the guidance in paragraph 88 of IAS 39.
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 14 of 22 |
The MCFH models incorporate various types of cash flows stemming from financial instruments that share LIBOR re-pricing risk. As illustrated in IAS 39.AG100, a hedging relationship can involve a hedge ratio of other than 1:1 in order to improve hedge effectiveness:
“If there is a valid statistical relationship between the two variables (i.e. between the unit prices of Brazilian coffee and Colombian coffee), the slope of the regression line can be used to establish the hedge ratio that will maximise expected effectiveness. For example, if the slope of the regression line is 1.02, a hedge ratio based on 0.98 quantities of hedged items to 1.00 quantities of the hedging instrument maximises expected effectiveness….”
The Swiss franc hedge accounting relationship within the MCFH model incorporates some limited hedge ratios of other than 1:1 in order to incorporate the cash flows of a Money Market Linked Account (“MMLA”) product which pays a percentage of the 1-month LIBOR interest rate. The relevant cash flows from this product have always been designated in the MCFH model taking into account a hedge ratio of other than 1:1. However, an operational update in Q4 2011 erroneously removed this ratio from our hedge effectiveness calculation, and this was identified as a result of analytical review work performed periodically to understand components of change within the models. As these MMLA cash flows constitute only a portion of the cash flows that are the hedged items in the model, this error did not change the assessment of hedge effectiveness in Q4 2011 and Q1 2012. However, it created an error in the ineffectiveness booked in Q4 2011 and Q1 2012; i.e. gains of CHF 23 million and CHF 5 million, respectively, that were corrected in the Q2 2012 Form 6-K.
Comment 17
Risk management and control – Credit risk, page 50
Exposure to selected European countries, page 54
We note your response to prior comment 30. Please tell us in more detail what information you are currently able to track and/or provide based on your current internal data and systems, and why specifically you are unable to disclose the information outlined in Topic 4 of the Division of Corporate Finance’s Disclosure Guidance: European Sovereign Debt Exposures. More specifically tell us whether you are able to quantify the fair values and notional values of credit default contracts purchased and sold by country and if so, revise your future filings to disclose such amounts. Please provide us with the disclosure you intend to provide in future filings.
For derivatives, our reported exposures include the counterparty risk (replacement value) against the country of domicile of the counterparty (presented within “Traded products”) as well as, where relevant, the issuer risk calculated as the instantaneous fall in value of the underlying reference asset to zero (assuming no recovery) that is shown against the country of domicile of the issuer of the reference asset (presented within “Trading inventory”).
The counterparty risk is reported from our credit risk system whereas the issuer risk is reported from our market risk system. For derivatives referencing a basket of instruments (e.g. credit indices), the issuer risk is decomposed into the individual issuers on a jump-to-zero basis. The issuer risk against each reference entity is calculated as the expected change in fair value of the derivative given an instantaneous fall in value to zero of the corresponding reference asset (or assets) issued by that entity. Issuer exposures are then aggregated across issuers (floored at zero per issuer) to reach an aggregate country figure. In contrast to the decomposition approach for issuer risk, the counterparty risk is reported as the replacement value against the single counterparty of the trade.
Notional and replacement values for derivatives are available within our credit risk system, but linking these to the decomposed issuer risk exposures within the market risk system is not currently possible in an automated and robust manner.
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 15 of 22 |
For single name credit default swaps, we can identify counterparty exposures where the reference asset is domiciled in a so-called “GIIPS” country (i.e. in any of Greece, Italy, Ireland, Portugal or Spain) and we can quantify the notional and fair value amounts of protection bought and sold per country on a gross and net basis. Furthermore, we can quantify the portion of this exposure where the counterparty is domiciled in a GIIPS country and specifically where the domicile of the reference asset is the same as the domicile of the counterparty (so called “risk-on-risk”).
In our Q4 2012 Form 6-K we enhanced our disclosure as follows:
“CDS are primarily bought and sold in relation to our trading businesses, but are also used to hedge parts of our risk exposure, including that related to selected eurozone countries. At 31 December 2012, and not taking into account the risk-reducing effect of master netting agreements, we had purchased approximately CHF 91 billion gross notional of single name CDS protection on issuers domiciled in Greece, Italy, Ireland, Portugal or Spain (“GIIPS”) and had sold CHF 88 billion gross notional of single name CDS protection. On a net basis, taking into account the risk reducing effect of master netting agreements, this equates to approximately CHF 18 billion notional purchased and CHF 15 billion notional sold. More than 99% of gross protection purchased was from investment grade counterparties (based on internal ratings) and on a collateralized basis. The vast majority of this was from financial institutions domiciled outside the eurozone. Less than CHF 1 billion of the gross protection purchased was from counterparties domiciled in a GIIPS country and less than CHF 0.5 billion was with counterparties domiciled in the same country as the reference entity.
Holding CDS for credit default protection does not necessarily protect the buyer of protection against losses, as the contracts will only pay out under certain scenarios. The effectiveness of our CDS protection as a hedge of default risk is influenced by a number of factors, including the contractual terms under which the CDS was written. Generally, only the occurrence of a credit event as defined by the CDS terms (which may include among other events, failure to pay, restructuring or bankruptcy) results in a payment under the purchased credit protection contracts. For CDS contracts on sovereign obligations, repudiation can also be deemed as a default event but would need legal clarification. The determination as to whether a credit event has occurred is made by the relevant International Swaps and Derivatives Association, Inc. (ISDA) determination committees (comprised of various ISDA member firms) based on the terms of the CDS and the facts and circumstances surrounding the event.”
Our proposed enhanced disclosure for the 2012 Form 20-F is as follows:
Exposure from Single Name Credit Default Swap referencing GIIPS domiciled entity
CHF millions
Protection Bought | ||||||||||||||||||||||||||||||||||||||||||||||||
o/w counterparty domiciled in | o/w counterparty domicile = | Net Position | ||||||||||||||||||||||||||||||||||||||||||||||
GIIPS country | reference entity domicile | Protection Sold | (after application of counterparty master netting agreements) | |||||||||||||||||||||||||||||||||||||||||||||
Notional | RV | Notional | RV | Notional | RV | Notional | RV | Buy Notional | Sell Notional | Positive RV | Negative RV | |||||||||||||||||||||||||||||||||||||
Greece | 1,405 | 155 | — | — | — | — | -1,388 | -162 | 346 | -329 | 54 | -62 | ||||||||||||||||||||||||||||||||||||
Italy | 47,884 | 2,285 | 550 | 26 | 226 | 3 | -46,406 | -2,460 | 8,024 | -6,394 | 365 | -539 | ||||||||||||||||||||||||||||||||||||
Ireland | 6,363 | 32 | 22 | 0 | — | — | -6,446 | -84 | 1,442 | -1,526 | 70 | -121 | ||||||||||||||||||||||||||||||||||||
Portugal | 7,163 | 387 | 71 | 9 | 7 | -0 | -7,110 | -430 | 1,702 | -1,622 | 124 | -167 | ||||||||||||||||||||||||||||||||||||
Spain | 27,702 | 968 | 289 | 6 | 129 | 4 | -26,994 | -990 | 6,355 | -5,395 | 320 | -342 | ||||||||||||||||||||||||||||||||||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |||||||||||||||||||||||||
Total | 90,516 | 3,828 | 933 | 42 | 362 | 7 | -88,343 | -4,126 | 17,869 | -15,266 | 933 | -1,231 | ||||||||||||||||||||||||||||||||||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Comment 18
As a related matter, your response to prior comment 30 indicates that your tabular disclosure on page 55 may include indexed or tranched purchased credit derivatives. Please revise your future filings to clearly disclose this fact, separately quantify the amount of this protection, and disclose how the amounts are reflected in your table amongst the different countries and counterparties.
We propose expanding our disclosure in our 2012 Form 20-F to more clearly explain how basket derivatives are reflected in our table as shown below, with additional wording underlined.
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 16 of 22 |
“In the case of derivatives, we show the counterparty risk associated with the positive replacement value against the country of domicile of the counterparty (presented within “Traded products”).
In addition, the risk associated with the instantaneous fall in value of the underlying reference asset to zero (assuming no recovery) is shown against the country of domicile of the issuer of the reference asset (presented within “Trading inventory”). This approach ensures that we capture both the counterparty and, where applicable, underlying issuer risk arising from derivatives and is applied comprehensively forall derivatives, including single-name CDS andother credit derivatives.
As a basic example: if a CDS protection for a notional value of 100 bought from a counterparty domiciled in country X referencing debt of an issuer domiciled in country Y has a positive replacement value of 20, we record (i) the fair value of the CDS (20) against country X (within “Traded products”) and (ii) the hedge benefit (notional minus fair value) of the CDS (100–20=80) against country Y (within “Trading inventory”). In the example of protection bought, the 80 hedge benefit would offset against any exposure arising from securities issued by the same entity as the reference asset, floored at zero per issuer. In the case of protection sold, this would be reflected as a risk exposure of 80 in addition to any exposure arising from securities held and issued by the same entity as the reference asset.In the case of derivatives referencing a basket of assets, the issuer risk against each reference entity is calculated as the expected change in fair value of the derivative given an instantaneous fall in value to zero of the corresponding reference asset (or assets) issued by that entity. Exposures are then aggregated by country across issuers, floored at zero per issuer.”
Comment 19
In your response to prior comment 31 you agreed to consider enhancements to your narrative discussion in future filings to address your process for monitoring and/or mitigating indirect exposures to selected European countries. At a minimum, we believe the information provided in your response would provide beneficial information to readers. Please provide us with your proposed disclosure revisions.
We intend to expand our disclosure in the 2012 Form 20-F as follows:
“In addition to monitoring direct exposure, we actively consider the inter-linkages among eurozone countries and institutions. We monitor and evaluate the policy responses of key EU institutions and the International Monetary Fund. In addition, we evaluate the implications of these developments for a broad range of countries and institutions beyond Europe when calibrating our eurozone-focused stress scenarios and making assumptions about the behavior of a variety of factors, including currencies, GDP, equity markets, consumer price index, corporate spreads, sovereign CDS and interest rates, for a number of key countries and regions. We apply these stress scenarios to our risk portfolios as part of our firm-wide stress testing framework. Furthermore, we subject our OTC exposures with a wide range of counterparties to these stress scenarios to gain an understanding of potential adverse impacts on our counterparty exposures, as well as to help identify so-called wrong-way risks.”
Comment 20
Interim consolidated financial statements
Notes to the interim consolidated financial statements, page 81
Note 12 – Fair value of financial instruments, page 89
Please address the following with respect to your response to prior comment 36:
• | In your response, you state that you believe your funds transfer price (FTP) curve is more appropriate than observed yields or spreads on your own traded notes for use in valuing your fair value option (FVO) liabilities provided those liabilities are issued at market clearing levels. Please clarify what you mean by “market clearing levels” and explain any changes made to your methodology in the event that such liabilities aren’t issued at market clearing levels. |
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 17 of 22 |
By “market clearing levels” we mean the spread at which market participants would be willing to trade UBS medium term notes (“MTNs”) and would be expected to regard them as being valued in line with similar products from other issuers in the market, taking into account differences of credit risk between issuing entities. As noted in our previous response, given the present state of the financial markets and our evolving response to them, which could include among other things a diminution in the volumes of MTNs we issue, it may be necessary in the future to refine the inputs to the funds transfer price (FTP) curve further or to use another measure outright in order to determine the fair value of UBS’s uncollateralized and partially uncollateralized funding transactions. We continue to monitor and assess the market situation and, were we to change our methodology, we would describe the rationale for the change in our future filings.
Comment 20, continued
• | In your disclosure on page 91 you state that the FTP curve spread is considered to be representative of the credit risk which reflects the premium that market participants require to purchase your medium term notes (MTNs). Please clarify how the FTP curve “spread” is being calculated in order to capture the premium required by market participants (e.g. what is the FTP curve being measured against to derive such a “spread”?). |
The starting point for the calculation of the FTP curve spread is the senior debt curve, which is generated from estimates of new market issuance levels for new senior unsecured UBS debt in sizes of at least USD 1 billion. These spreads are then adjusted in order to arrive at the FTP curve, with the adjustment reflecting the structured and bespoke nature of MTN issuances, the reverse enquiry nature of the MTN market, and the smaller issuance sizes involved.
Comment 20, continued
• | Tell us how you considered the guidance in paragraphs AG74-AG76 of IAS 39 in determining that your FTP curve is an input that would be used by market participants in pricing these instruments given that this curve appears to represent an entity-specific cost of funding. Further, given that the FTP curve is based on internal funding rates, tell us how you determined that it reflects your true cost of external funding and does not include any internal treasury funding incentives. |
Liabilities reported as held at fair value largely consist of structured notes for which the secondary market is limited due to their bespoke nature and small issuance sizes. It is therefore not possible to directly observe independent fair values for these positions. As such, UBS relies on valuation techniques to establish prices. For UBS to exit structured notes, the price it would have to pay is the cost to source replacement funding. The FTP curve incorporates the primary factors that market participants would consider in valuing our debt instruments. The methodology used to construct our FTP curve maximizes the use of available market inputs (secondary trading in vanilla UBS debt, primary issuance premium, and other independent market based data) and is periodically reviewed against our actual issuance levels. As described above, the FTP curve is set with reference to “market clearing levels”, and explicitly does not include any internal treasury funding incentives.
Comment 20, continued
• | Clarify how you evaluate the potential for overlap of certain risks between the FTP curve and other measures of your own credit risk. If you determine that an overlapping of such risks exists, explain whether you make any adjustments to avoid potentially double-counting those risks. |
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 18 of 22 |
Fair value option (“FVO”) liabilities are discounted using the FTP curve for valuation purposes and no other measures of our own credit risk are included in these valuations. Population controls are in place to ensure that there is no duplication between these FVO liabilities and the derivative population, where our own credit risk is incorporated into the valuation using a debit valuation adjustment (DVA). There are also controls in place to ensure that no trades in the derivative population are discounted using the FTP curve in addition to having a DVA applied to them.
Comment 20, continued
• | Clarify whether you make any additional adjustments to the FTP curve used to discount your fair value option liabilities. If so, explain the reasons for such adjustments and clarify whether they were made as aggregate adjustments to the fair value of such liabilities or as incremental adjustments to the spread. |
As the FTP spreads are constructed from underlying MTN pricing curves, some curve smoothing adjustments take place between tenors in order to ensure that the resulting curve produces forward rates that are sufficiently smooth to be suitable for the discounting of callable instruments. The valuation impact of this curve construction approach on the FVO liability population is quantified monthly and to date has always been at a level which is considered immaterial. If the impact were to become material in any given period, an aggregate adjustment would be made to the fair value of the population. No other aggregate adjustments are made.
Comment 21
Form 6-K filed October 30, 2012
Notes to the interim consolidated financial statements, page 75
Note 17 – Provisions and contingent liabilities, page 90
b) Litigation, regulatory and similar matters – 6. Claims related to sales of residential mortgage-backed securities and mortgages – Loan repurchase demands related to sales of mortgages and RMBS, page 93
We note your response to prior comment 37. Please address the following:
• | We note that the number of loan repurchase demands by year received included in the line item “demands in litigation” increased significantly (94%) between June 30, 2012 and September 30, 2012. Please expand your response to bullet point one to tell us in more detail, and disclose in future filings, the specific reason(s) you do not expect to make payment for most of the loan repurchase demands received during the periods that are currently in litigation. In this regard, clarify whether your determination is based on a loan-by-loan review or historical claims experience. Please provide us with your proposed disclosure. |
By way of background, UBS RESI was sued by Assured Guaranty (“AGO”) in February 2012 in connection with approximately USD 997 million in original principal balance (“OPB”) of loans for which AGO had previously demanded repurchase (the “AGO Suit”). At the time the AGO Suit was filed, this amount represented most, but not all, of the repurchase demands made by AGO since 2010. The repurchase demands at issue in the AGO Suit represented the vast majority of the USD 1,091.5 million of “demands in litigation” as of June 30, 2012 (the remaining approximately USD 94 million reflected repurchase demands made by Freddie Mac that are the subject of a separate litigation filed in April 2012).
As described in the Q3 2012 Form 6-K disclosures, the Court ruled on UBS’s motion to dismiss the AGO Suit in August 2012, and held that AGO could not assert claims seeking to enforce UBS’s obligation to repurchase loans, and that only the trustee for the RMBS trusts at issue could enforce those claims. Consistent with the Court’s ruling, and as disclosed, the RMBS trusts filed suit in October 2012 seeking to enforce UBS’s obligation to repurchase loans that had previously been put back by AGO (the “Trustee Suit”). However, the Trustee Suit seeks to enforce repurchase demands for approximately USD 2 billion in OPB of loans put back by AGO (this amount represents most, but not all, of the repurchase demands made by AGO since 2010, and includes the USD 997 million of demands that were at issue in the AGO Suit and additional repurchase demands made after the filing of the AGO Suit). Accordingly, the filing of Trustee Suit accounts for the 94% increase in the “demands in litigation” category in the Q3 2012 Form 6-K disclosures.
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 19 of 22 |
Based on our internal review, the approximately USD 2 billion in OPB of loans currently at issue in the Trustee suit were sold to UBS principally by five originators:
• | Countrywide (approximately USD 1.13 billion) |
• | Mortgage IT (approximately USD 61 million) |
• | American Home Mortgage (approximately USD 745.6 million) |
• | RFC (approximately USD 55 million) |
• | IndyMac (approximately USD 30 million) |
The latter three originators are insolvent and are unable to provide indemnity to UBS RESI for the repurchase demands at issue in the Trustee Suit, as required under the terms of the mortgage loan purchase agreement (“MLPA”) pursuant to which UBS RESI acquired the loans. Accordingly, UBS has undertaken to perform a loan-by-loan review for loans from these originators where it has been able to access the loan files from the servicer (approximately 95% of such loans from the latter three originators have been reviewed to date). UBS has provisioned for defaulted loans from these originators based on current loss severities for the loans and UBS’s observed rate of concurrence with the allegations of breach of representation and warranty (“R&W”) as a result of UBS’s review of loan files. Countrywide remains in business and is available to indemnify UBS for the loans that it originated and that are now the subject of the Trustee Suit. Similarly, indemnification is available from Mortgage IT, which was acquired by Deutsche Bank. The indemnification obligations in the MLPAs with these originators are broad and extend to “any liability” arising from a breach of R&W concerning the mortgage loans. UBS will bring legal action against these originators to enforce its indemnity rights should the parties not be able to reach agreement on the indemnity owed to UBS. UBS has communicated its position to these originators. At the appropriate time, such an action would be brought together with the AGO Suit and the Trustee Suit. After consultation with external counsel, we believe that UBS would be likely to prevail against these originators in an action for indemnification should the Countrywide or Mortgage IT loans be found to be in breach of R&W.2
We propose including enhanced disclosure in the 2012 Form 20-F as follows:
“At this time, UBS does not expect that it will be required to make payment for the majority of loan repurchase demands at issue in the suit brought by the RMBS trusts for at least the following reasons: (1) we reviewed the origination file and/or servicing records for the loan and concluded that the allegations of breach of representations and warranties are unfounded, or (2) a surviving originator is contractually liable for any breaches of representations and warranties with respect to loans that it originated. UBS has indemnification rights in connection with approximately half of the USD 2 billion in original principal balance of loans at issue in this suit (reflected in the “In litigation” category on the accompanying table). Additionally, in its motion to dismiss the suit filed by the Trusts, UBS has asserted that, under governing transaction documents, UBS is not required to repurchase liquidated loans that were the subject of repurchase demands now at issue in this suit.”
2 | We note that a peer bank has similarly pursued enforcement of its indemnification rights by impleading a loan originator in a lawsuit seeking to enforce repurchase demands and that this approach has been sanctioned by the courts.See Decision and Order dated 7/25/11 inAssured Guaranty v. DB Structured Products(denying motion to dismiss defendant’s third-party complaint for indemnification against GreenPoint Mortgage Funding, Inc., the originator of the loans put back by plaintiff, and holding that “[n]o matter how strictly the indemnification provision is construed, if certain allegations in Assured’s complaint are true, defendant is entitled to indemnification from GreenPoint”). |
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 20 of 22 |
Comment 21, continued
• | We also note that loan demands resolved or expected to be resolved through enforcement of indemnification rights against third-party originators decreased significantly (-81%) between June 30, 2012 and September 30, 2012. Tell us the reason(s) for this decrease, and clarify whether these demands were resolved or reclassified into a different line item within your table. Additionally, tell us the success rate you have had with collecting on your indemnification rights against surviving third-party originators. To that effect, please provide support for your assertion that you do not provision for loan repurchase demands for which you seek to enforce indemnification rights because you do not expect to experience an outflow of resources and tell us how you considered the guidance in paragraphs 53-58 of IAS 37. |
The “demands resolved or expected to be resolved through enforcement of indemnification rights against third-party originators” decreased for the same reasons described above. That is, a significant number of loans put back to UBS by AGO that were originated by Countrywide and Mortgage IT (and for which UBS intends to seek indemnity) were not included in the AGO Suit, but were included in the Trustee Suit. As a result, when the Trustee suit was filed in early October 2012, these loans were reclassified from the “demands resolved or expected to be resolved through enforcement of indemnification rights against third-party originators” category to the “demands in litigation” category.
We have not yet enforced our indemnification rights against surviving third-party originators. Please see the response to bullet point one, above, for UBS’s expectations regarding the enforcement of its indemnification rights.
In regards to IAS 37.53-58, we note that third-party indemnitors sold mortgage loans to UBS pursuant to various sale/purchase agreements (MLPAs) which include the obligation to indemnify UBS for any liability arising from representations and warranties made with respect to the originated loans. Accordingly, UBS will bring legal action against those third-party indemnitors, asserting that they have ultimate liability as both the originators of the loans and pursuant to contractual indemnification obligations. As previously indicated a peer bank has similarly pursued enforcement of its indemnification rights by impleading a loan originator in a lawsuit seeking to enforce repurchase demands and this approach has been sanctioned by the courts. From an accounting perspective, we also believe that the approach shows that UBS’s obligation is limited to only that which the third-party indemnitors will not cover. Thus, we believe that these facts and circumstances are functionally equivalent to those described in IAS 37.58, as failure to pay could occur only as a result of bankruptcy similar to a scenario of joint and several liability.
Furthermore, we believe that this situation is substantially different from a standard reimbursement agreement as the third-party indemnitors do not have a reimbursement agreement but rather a contractual obligation to indemnify. Furthermore, as opposed to a standard financial guarantee, insurance or reinsurance arrangement, whereby the reimbursing party has agreed to make payments upon the occurrence of insured events in exchange for premiums, in this case, the third-party indemnitors have originated the assets giving rise to the associated liabilities. Consequently, UBS is effectively acting as a go-between with respect to AGO and the third-party indemnitors, who bear ultimate liability for claims of breach of representations and warranties on the underlying loans. As such, UBS’s obligation is that which is not expected to be covered by the third-party indemnitors.
Comment 21, continued
• | On a related note, we note that demands rebutted by you but not yet rescinded by counterparty decreased 75% between June 30, 2012 and September 30, 2012. Please tell us the reason(s) for this decrease, and whether those demands were formally rescinded or reclassified into a different line item within your table. |
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 21 of 22 |
The “demands rebutted but not yet rescinded” decreased for the same reasons described above in the response to bullet point one. That is, a significant number of loans that had been reviewed by UBS and rebutted but not rescinded by AGO were not included in the AGO Suit, but were included in the Trustee Suit. As a result, when the Trustee suit was filed in early October 2012, these loans were reclassified from the “demands rebutted but not yet rescinded” category to the “demands in litigation” category.
Comment 21, continued
• | Based on your response to bullet point number three, it appears that you have not provisioned for any of the repurchase demands received from Assured Guaranty (AGO). Please confirm whether this is true and, if so, tell us in more detail why you believe you should not record a provision related to the approximately USD 513.8 million of “demands rebutted by UBS but not yet rescinded by counterparty” (e.g. why do you believe the counterparty will rescind these claims in full?). |
UBS has, in fact, recognized a provision for many of the loan repurchase demands made by AGO. UBS has currently set a provision for defaulted loans put back by AGO where (i) there is no surviving originator to provide indemnity, and (ii) where UBS’s review of the loan files resulted in a concurrence with the allegations of breach of R&W. UBS has not provisioned for loans that are not in default, or that were rebutted after review of the loan file or otherwise rebutted for failure to allege breach of R&W. This provision totals approximately USD 113 million.
UBS believes that many of the loans for which AGO seeks repurchase are not properly the subject of a repurchase demand and that UBS ultimately will not face repurchase liability for such loans. For example, of the USD 483.1 million in OPB of loans in the “rebutted but not yet rescinded by counterparty” category in UBS’s Q3 2012 Form 6-K disclosures, approximately 60% of this amount relates to loans put back by AGO that were securitized in the MABS 07-NCW trust. But the allegations of breach of R&W propounded by AGO do not correspond to the narrow and limited R&W in the deal documents, and the loan repurchase demands made by AGO from late December 2011 to mid-December 2012 were accordingly rebutted. AGO has not responded to or challenged UBS’s rebuttals of these demands, nor are these demands included in the AGO Suit or the Trustee Suit. Accordingly, UBS does not believe that these repurchase demands will ultimately result in liability.
Similarly, the remaining 40% of the demands in the “rebutted but not rescinded” category were rebutted following a detailed review of the loan file, and UBS believes that the allegations of breach are without merit and should ultimately be rescinded.
Comment 21, continued
• | We note based on your response that the vast majority of repurchase demands received to-date are from AGO. To the extent that you begin to see an increase in demands from other counterparties, we believe that additional disclosure regarding your claims experience by type of counterparty would be beneficial to readers. Please revise future filings accordingly. |
We will enhance our disclosure to the extent that such increase in demands from other counterparties occurs.
* * * * * * *
UBS AG Form 20-F for the Fiscal Year Ended December 31, 2011 filed March 15, 2012 Forms 6-K filed July 31, 2012 and October 30, 2012 Response dated September 10, 2012 File No. 001-15060 28 February 2013 Page 22 of 22 |
We appreciate the opportunity to respond to your Comments. If you wish to discuss any of our responses or require additional information, please do not hesitate to contact either Todd Tuckner at +1 203 719 0381 (todd.tuckner@ubs.com) or Alex Brougham at +44 207 567 7425 (alex.brougham@ubs.com).
Very truly yours,
UBS AG | ||
/s/ Tom Naratil
Tom Naratil | /s/ Todd Tuckner
Todd Tuckner | |
Group CFO | Group Finance COO |