FOIA Confidential Treatment Requested by Citigroup Inc.

John C. Gerspach

Chief Financial Officer

Citigroup Inc.

399 Park Avenue

New York, NY 10022

September 21, 2012

VIA EDGAR CORRESPONDENCE

Stephanie J. Ciboroski

Securities and Exchange Commission

100 F Street, NE

Mail Stop 4561

Washington, DC 20549-7553

Re: Citigroup Inc.

Form 10-K for Fiscal Year Ended December 31, 2011

Filed February 24, 2012

Form 10-Q for the Quarterly Period Ended June 30, 2012

Filed August 3, 2012

Response filed April 27, 2012

File No. 001-09924

Dear Ms. Ciboroski:

Enhancement of Citigroup’s disclosures is an objective that we share with the staff (the “Staff”) of the Securities and Exchange Commission (the “SEC”) and one that we consider in all of our filings. This letter sets forth the responses of Citigroup to the comments of the Staff contained in the Staff’s letter dated August 22, 2012.

For your convenience, the text of the Staff’s comments is set forth in bold below, followed in each case by our response.

Form 10-Q for the Quarterly Period Ended June 30, 2012

Capital and Loan Loss Reserve Positions, page 6

1.We note your disclosure of estimated Tier 1 Common ratio under Basel III at the end of the second quarter of 2012 here as well as on page 40. Additionally, we note your disclosure on page 43 of your estimated Tier 1 Common ratio as of June 30, 2012, assuming application of the final Basel II.5 rules. These measures appear to represent non-GAAP measures as they are not yet a required disclosure by a government, governmental authority or self-regulatory agency. Therefore, please expand your disclosures in future filings to show the calculation of the metrics, similar to the disclosures on page 40 where you show the calculation of the metrics under current regulatory requirements.

To the extent we include these metrics in our future filings, we will expand our disclosure to indicate that (i) Citi’s estimated Basel III Tier 1 Common ratio and Citi’s estimated Basel I Tier 1 Common ratio, assuming

CITI B-1

application of the final Basel II.5 market risk rules, are non-GAAP financial measures, (ii) why management believes the measures provide useful information to investors and (iii) the calculation of each such metric, similar to our current “Components of Capital Under Current Regulatory Guidelines” table on p. 40 of our Second Quarter 2012 Form 10-Q (“2Q 2012 Form 10-Q”) or a similar method of disclosure to show the calculation of these metrics.  We supplementally advise the Staff that while Citi’s Basel III Tier 1 Common ratio may not technically be “required disclosure” by a government, governmental authority or self-regulatory agency, we believe our investors and other stakeholders are very interested in this information.  We further advise the Staff that the final Basel II.5 market risk rules are to become effective on January 1, 2013 and, as such, will be fully reflected in Citi’s risk-based capital ratios that will be reported commencing with the first quarter of 2013.  However, the U.S. Basel III risk-weighted assets and capital rules are not yet final and therefore we will continue to disclose the estimated Basel III Tier 1 Common ratios as a non-GAAP measure.

Segment and Business — Income (Loss) and Revenues, page 11

Citi Holdings — Brokerage and Asset Management, page 29

MSSB JV, page 29

2.We note your response to prior comment 7 from our letter dated April 9, 2012, as well as your updated disclosure regarding new developments in the second quarter related to the MSSB JV Agreement deal with Morgan Stanley. Due to the fact that you could potentially record a significant non-cash charge in the third quarter of 2012 related to this disposition, which would be based in part on your internal valuation of the fair value of that portion of your investment, please provide us with a copy of your valuation report used in determining the fair value of your interest in the MSSB JV for purposes of this transaction.

As disclosed pursuant to a Form 8-K filed with the SEC on September 11, 2012, Citi anticipates recording a non-cash charge to net income of approximately $2.9 billion after-tax ($4.7 billion pretax) in the third quarter of 2012 relating to its equity method investment in the Morgan Stanley Smith Barney joint venture. We have provided to the Staff the information it has requested under separate cover as supplemental information pursuant to Rule 12b-4 under the Securities Exchange Act of 1934.

Corporate/Other, page 34

3.We note your disclosure on page 34 that your net loss in the Corporate/Other segment for 2Q11 and 2Q12 was $63 million and $437 million, respectively. However, we also note your statement that the net loss of $437 million “declined by $374 million year-over-year” implying that the net loss decreased when in fact it grew to a larger deficit. Similar disclosure is included in the year to date comparison on the same page. In the future, please ensure that your disclosure accurately depicts the changes between each period discussed.

In future filings, as applicable, we will clarify our disclosures as requested by the Staff.

Liquidity Management and Measures, page 49

Liquidity Measures, page 49

4.We note your disclosure that you believe you are in compliance with the proposed Basel III Liquidity Coverage Ratio (LCR) with an estimated LCR of approximately 117% as of June 30, 2012, compared with approximately 126% at March 31, 2012. Given that this metric is not yet required by a government, governmental authority or self-regulatory organization, it would appear to be a non-GAAP financial measure for purposes of Item 10 of Regulation S-K. Thus, please revise

CITI B-2

future filings to label the measure as non-GAAP and provide further clarifying disclosures regarding the use of the metric, including more discussion on how it is calculated.

To the extent we include this metric in our future filings, we will expand our disclosure to indicate that (i) Citi’s estimated LCR is a non-GAAP financial measure, (ii) why management believes the measure provides useful information to investors and (iii) the calculation of such metric.  We supplementally advise the Staff that while Citi’s LCR may not technically be “required disclosure” by a government, governmental authority or self-regulatory agency, given that this metric has been proposed by the Basel Committee and U.S. banking regulators and is part of the significant proposed changes to capital and liquidity standards applicable to financial institutions, we believe our investors and other stakeholders are very interested in this information.

Managing Global Risk, page 52

Credit Risk — North America Consumer Mortgage Lending, page 59

North America Residential First Mortgages — State Delinquency Trends, page 61

5.We note your table showing significantly improved loan-to-value (LTV) percentages between March 31, 2012 and June 30, 2012 for certain states and state groupings, including California, Florida, Arizona/Nevada and Other. Please tell us and expand your disclosure as appropriate to discuss the primary drivers of the significant improvements in LTVs over such a short period (for example, 20% fewer loans with LTVs greater than 100% in California from one quarter to the next). To the extent that loans are concentrated in certain geographic regions within the state(s), please state that fact. To the extent that modifications that include principal forgiveness are a key driver, please state that fact and reconcile the information with the disclosures on page 143.

The improvement in LTV percentages from March 31, 2012 to June 30, 2012 in total and in California, Florida, Arizona/Nevada and Other states was primarily the result of improvements in the home pricing index (HPI) across substantially all metropolitan statistical areas, thereby increasing the house value (denominator) which reduced the LTV to below 100% on a portion of loans that had LTVs greater than 100% as of March 31, 2012.  While principal forgiveness during the second quarter had an impact on the improved LTV, it was not a key driver of the improvement.  In future filings, we will expand our disclosures to discuss the primary drivers of any significant changes in these percentages, including any significant geographic concentrations within the state and/or regions that are contributing to the changes, to the extent material.

North America Consumer Mortgage Quarterly Credit Trends — Delinquencies and Net Credit Losses — Home Equity Loans, page 62

6.We note your disclosure that approximately $23.5 billion of home equity lines of credit are still within their revolving period and have not commenced amortization, and that this amount has decreased from $27.0 billion as of December 31, 2010. We also note your response to our prior comment 7 in your letter dated June 29, 2011 that a comparison of default and delinquency trends for amortizing versus non-amortizing home equity lines of credit was not meaningful at that time given the levels of loans that had converted to amortizing as of March 31, 2011. However, given the time that has passed, and the fact that more loans have now converted to amortizing, please tell us and expand your disclosures as appropriate to tell us whether there are different trends for the loans that have converted to amortizing versus those that are still in the revolving stage.

As noted in the Staff’s comment, the amount of home equity lines of credit in their revolving period (which have not commenced amortizing) has declined approximately $3.5 billion since December 31, 2010.  The decline during this period was primarily a result of the home equity lines of credit paying off or charging off and not a result of the conversion of these lines from revolving to amortizing.  The outstanding balance of home equity lines of credit that have converted from revolving to amortizing remains at a low level consistent

CITI B-3

with the contractual conversion schedule provided in our response letter dated June 29, 2011 and as we have disclosed in our quarterly and annual filings.  Accordingly, we continue to believe that disclosure of the comparison of default and delinquency trends between amortizing and revolving home equity loans is not material at this time; however, management does and will continue to monitor these default and delinquency trends and, to the extent material, we will expand our disclosures in future filings to separately present default and delinquency statistics for amortizing and non-amortizing home equity lines of credit.

Representation and Warranty Claims — by Claimant, page 70

7.We note your disclosure on page 69 that you refined certain of your assumptions in estimating the repurchase reserve. Specifically, you now use the historical correlation between underlying loan characteristics (e.g. delinquencies, LTV, loan channel, etc.) and the likelihood of receiving a claim based on those characteristics in developing your claim estimate. Prior to this change, you used loan documentation requests and repurchase claims as a percentage of loan documentation requests in order to estimate future claims. Your disclosure also states that the refinement did not have a material impact on the repurchase reserve balance at June 30, 2012. Please tell us how you believe this new approach captures the changes in GSE behavior regarding the levels of mortgage repurchase requests, as noted by many other financial institutions, and as evidenced by the higher levels of GSE mortgage repurchase claims you experienced during the first and second quarters of 2012. In this regard, it appears you are now experiencing some of the highest levels of GSE repurchase requests you have ever experienced, despite the fact that you disclose on page 68 that the repurchase requests are primarily related to loans originated from 2006 to 2008. We also note that in June 2012, the FHFA indicated they would provide a framework to assist banks in developing a higher degree of certainty and clarity around repurchase exposure and liability, as well as consistency around repurchase timelines, incentives and remedies. Please tell us how this was taken into consideration in making the policy change and tell us whether you have received any guidance at this point. Finally, in light of your change in methodology where you use historical correlation based on underlying loan characteristics, tell us why you still believe that the activity in, and change in, estimate for the repurchase reserve will remain volatile, and that you cannot estimate a range of reasonably possible loss related to this exposure.

As we disclose, Citi’s repurchase reserve primarily relates to whole loan sales to the GSEs and continues to be calculated primarily based on Citi’s historical repurchase activity with the GSEs.  We believe our refined approach to determining future claims for purposes of the whole loan repurchase reserve continues to capture any changes in GSE behavior regarding documentation package requests since it is based on loan level characteristics, which we believe ultimately drive the potential for future claims, and is no longer influenced by factors such as the timing of documentation package requests. The refined approach also envisions periodically updating the historical correlation between underlying loan level characteristics and claims, therefore ensuring that it continues to consider GSE behavior regarding whole loan mortgage repurchase requests.  We advise the Staff that the largest portion of GSE claims continues to come from loans originated from 2006 to 2008, consistent with our disclosure.

Regarding the framework from the FHFA, we did not receive any direct communication from the FHFA until September 11, 2012.  Accordingly, no assumptions had been changed as of June 30, 2012 to account for any anticipated guidance. While the FHFA, Fannie Mae, and Freddie Mac announced a new representation and warranty framework on September 11, 2012, it will apply prospectively to loans sold to Fannie Mae or Freddie Mac on or after January 1, 2013. Therefore, any impact of this guidance will be reflected in our whole loan repurchase reserve for loans sold to the GSEs beginning with the first quarter of 2013.

Because the estimation process for this reserve includes multiple assumptions that are not driven by observable factors (for example, the reserve is highly dependent upon the environment and focus by the GSEs), we continue to believe that the activity in, and change in estimate relating to, Citi’s whole loan repurchase reserve will remain volatile in the near term, as we disclose.  However, in future filings, so long as

CITI B-4

material, we will disclose an estimate of the possible loss or range of loss related to whole loan sales within our repurchase reserve.

Market Risk, page 78

Value at Risk for Trading Portfolios, page 79

8.We note your discussion of your value at risk (VAR) methodology which is estimated using a Monte Carlo simulation. Please respond to the following and expand your disclosure in future filings as appropriate:

·Tell us the number of different VAR models that are used to determine your total trading VAR, and discuss the drivers regarding the need to use multiple different models.

·Tell us how all of the different VAR models used are aggregated to arrive at your total trading VAR as well as the individual market risk categories disclosed. For example, clarify whether you simply aggregate the outputs from the different models or whether adjustments are made, and if so, how the adjustments are determined.

We advise the Staff that Citi uses a single, coherent VAR model for all positions and asset classes/risk types, including individual market risk categories, to determine its total trading VAR, and the model parameters are consistent across all risk types. We believe our disclosure sets forth the key attributes of Citi’s VAR model:  Monte Carlo simulation, volatilities implied by the greater of the last month or last three years, 99% confidence level, and “fat tail” scaling.

·We note your disclosure indicates that the conservative features of the VAR calibration contribute approximately 15% add-on to what would be a VAR estimated under the assumption of stable and perfectly normally distributed markets, but that this add-on was 20% as of December 31, 2011 and March 31, 2012. Please discuss the drivers of the change in this estimated add-on and whether you expect it to change over time.

The decrease in the size of the add-on from fourth quarter of 2011/first quarter of 2012 to the second quarter of 2012 resulted from underlying market data related to the more systematically unstable market conditions of the financial crisis period being excluded from the three-year volatility time series used in Citi’s VAR calculation (see our disclosure on p. 79 of the 2Q 2012 Form 10-Q). In other words, as more time passes since the financial crisis period, the underlying market data from that unstable period is no longer included in the three-year volatility time series used in Citi’s VAR.  The above noted add-on reduction would be expected as the size of the add-on is inversely related to the level of normal market distribution. If market returns were normally distributed, the add-on would be nonexistent; if the market data were not normally distributed and showed extreme and abrupt variations, the add-on would increase substantially.

·We note your disclosure in footnote 4 that the credit portfolio is composed of the asset side of the CVA derivative exposures, all CVA hedges, and hedges to the loan portfolio, fair value option loans and tail hedges that are not explicitly hedging the trading book. Please tell us whether there are other hedges of other assets that are not explicitly included in this category, or specific securities portfolio that are held for investment rather than trading and give examples and quantification of amounts excluded to the extent possible.

We advise the Staff that no other economic or macro hedges of other assets are excluded from the category covered by footnote 4. Citi’s VAR model includes only positions that are measured at fair value with changes in fair value recognized in the Consolidated Statement of Income. It does not include investment securities classified as available-for-sale (“AFS”) or held-to-maturity (“HTM”). Information on the size of the AFS and HTM investment securities portfolios can be found in Note 11 of Citi’s 2Q 2012 Form 10-Q.

CITI B-5

·Discuss the type of risks excluded from your VAR calculation and the reasons why such risks are excluded.

We advise the Staff that Citi believes its VAR model and calculations capture all material risk sensitivities (such as first and second order sensitivities of positions to changes in market prices) of various risk types (such as interest rate, foreign exchange, equity and commodity risks).

·Tell us whether the VAR models used for regulatory capital purposes are the same as the VAR models used for your market risk disclosures. To the extent that certain of the models used for both purposes differ, please tell us the drivers behind those differences.

We advise the Staff that Citi uses the same VAR model for both regulatory capital and external market risk disclosure purposes. While the scope of positions included in the VAR model calculations for regulatory capital purposes differs from the scope of positions included in the VAR model calculations for external market risk disclosure purposes (e.g., the interest rate sensitivity of repos and reverse repos and the credit and market sensitivities of the derivatives CVA are included for external market risk disclosure purposes, but are not included for regulatory capital purposes), these differences are simply due to the fact that such positions are not eligible for market risk treatment under the U.S. regulatory capital rules, either as currently in effect under Basel I or as under the final Basel II.5 rules becoming effective January 1, 2013. The applicability of the VAR model for positions eligible for market risk regulatory capital treatment is periodically reviewed and approved by our primary U.S. banking regulators, the Federal Reserve and the OCC.

·Discuss the process and validation procedures in place prior to implementing significant model and assumption changes. For example, discuss the approval process required, back-testing procedures performed, and periods of parallel model runs before implementation.

We advise the Staff that significant VAR model and assumption changes must be independently validated within Citi’s risk management organization. This validation process includes a review by Citi’s model validation group and further approval from our model validation review committee which is composed of senior quantitative risk management officers. In addition, significant model and assumption changes are subject to periodic reviews and approval by our primary U.S. banking regulators. Furthermore, as part of our business as usual process, back-testing is performed against the actual change in market value of transactions on a quarterly basis at multiple levels of the organization (trading desk, ICG business segment and Citigroup levels), and the results are shared with our primary U.S. banking regulators. In the event of significant model changes, parallel model runs are undertaken prior to implementation, which follow prescriptive Citi policy requirements, with the testing period varying from 7 days for simple changes to 12 months for more complex changes.

Generally, Citi’s VAR model validation process entails reviewing the conceptual soundness of the modeling framework, the major assumptions and implementation of the mathematical algorithm. As part of this model validation process, product specific back-testing on hypothetical portfolios is periodically completed and reviewed with our primary U.S. banking regulators.

·To the extent that all or some of your VAR models used for your market risk disclosures are different than those used to calculate regulatory capital, please tell us whether the model review process and model oversight processes are the same for both. As part of your response, please clarify when approval is required from any of your regulators regarding VAR model changes.

As stated above, Citi uses the same independently validated VAR model for both regulatory capital and external market risk disclosure purposes (subject to the scope exceptions described above) and, as such, the model review and oversight processes for both purposes is as described above. As we note above, the

CITI B-6

applicability of the VAR model for positions eligible for market risk regulatory capital treatment is reviewed and subject to approval by our primary U.S. banking regulators.

·Given that your VAR models evolve over time, tell us how you consider when disclosure is required under Item 305(a)(4) of Regulation S-K regarding model, assumptions and parameter changes.

As the Staff is aware, Citi has previously disclosed significant changes in its VAR model calculation methodology (e.g., the implementation of exponentially weighted market factor volatilities for interest rate and FX positions to the VAR calculation, effective for the quarter ended June 30, 2010 (see p. 75 of Citi’s Form 10-Q for the second quarter of 2010)) as well as changes in the scope of positions considered (e.g., the inclusion for external market risk disclosure purposes of derivative CVA (see p.  98 of Citi’s Form 10-K for the fiscal year ended December 31, 2011 (the “2011 Form 10-K”)). Consistent with Item 305(a)(4) of Regulation S-K, Citi considers the materiality of the model, assumptions or parameter changes in determining whether disclosure of such changes is required.  Specifically, on a quarterly basis, senior management reviews any changes in the VAR model, assumptions, and/or parameters to determine whether such changes are material and thus require disclosure.  Conversely, regular parameter changes — such as the declining higher volatilities in the three-year volatility time series used in Citi’s VAR model calculation, as discussed above and disclosed — would not necessarily require disclosure, but Citi may include such disclosure if it believes such information is useful to the reader.

Country Risk, page 89

Overview, page 89

9.We note your response to prior comment 5 from our letter dated April 9, 2012, related to the last bullet point in that comment regarding indirect risk exposure. We believe indirect risk exposure is interpreted, defined and managed differently among companies and therefore believe additional clarifying disclosure outlining identified indirect exposures, along with quantification to the extent possible, increases transparency to investors. Examples of indirect risk exposures could include exposure to other entities that are highly exposed to the affected areas and counterparties, indirect effects resulting from the impact on Western European banking sectors, slowdown in global economic activity and wider economic contraction in the corporate sector, potential for redenomination and devaluation risk (which you have now separately discussed in detail beginning in your June 30, 2012 Form 10-Q), and related effects to payment and clearing systems or global exchanges. Therefore, please expand your disclosure to describe how you consider these factors, identify the related risks and manage them as part of your overall risk management process.

As noted in our prior response, Citi does not manage its risks related to country exposures pursuant to buckets of “direct” vs. “indirect” exposures.  Rather, Citi identifies and manages risks of loss related to its country exposures as part of the overall risk management process.  We believe we have identified and in fact disclosed many of the examples of what the Staff refers to as “indirect” risks in Citi’s filings, as well as how we manage such risks, although we do not expressly refer to them as “indirect” exposures.

For example, in the “Risk Factors” section of the 2011 Form 10-K (pps. 59-61), Citi noted the following examples of exposures that the Staff refers to as “indirect” risks:

·Disruptions in the global financial markets generally and general decline in global financial conditions (which could impact Citi’s overall results of operations);

·European banks and insurers experiencing widening credit spreads and the resulting decreased availability and increased costs of funding (which could impact Citi’s spreads as well as those of other global financial institutions);

CITI B-7

·Severe disruption to equity markets, significant increases in bond yields generally, potential failure or default of financial institutions (including those of systemic importance), a significant decrease in global liquidity, a freeze-up of global credit markets and worldwide recession;

·Exchange controls, limitations on foreign investment, socio-political instability, nationalization, closure of branches or subsidiaries, confiscation of assets and sovereign volatility; and

·How the ongoing Eurozone debt crisis could affect emerging markets; specifically, how a recession in the Eurozone could cause a ripple effect in emerging markets (which could disproportionately impact Citi in light of its emphasis on emerging markets as part of its global strategy).

In the “Managing Global Risk — Risk Management — Overview” section of the 2011 Form 10-K (pps. 66-67), Citi described how risk is managed across the company and noted, among other things:

·Citi has a comprehensive risk management process to monitor, evaluate, and manage the principal risks it assumes in conducting its activities, including credit, market and operational risks.

·The risk management organization is structured so as to facilitate the management of risk across three dimensions — businesses, regions and critical products — with additional details as to the organization and management within each dimension.

·The Chief Risk Officer monitors and controls major risk exposures and concentrations across the organization, including aggregating risks, within and across businesses, as well as subjecting those risks to alternative stress scenarios in order to assess the potential economic impact they may have on Citi.

·In addition to the stress testing described above, Citi independent risk management provides periodic updates to senior management on significant potential areas of concern across Citi that can arise from risk concentrations, financial market participants, and other systemic issues.

·In addition to enhancing awareness and understanding of potential exposures, the results of these processes then serve as the starting point for developing risk management and mitigation strategies.

We advise the Staff that this overall risk management process, as described in the 2011 Form 10-K and incorporated into subsequent periodic reports, is the process by which Citi identifies and manages the risks associated with its country risk exposures — including those relating to the Eurozone — as well as the other credit, market and operational risks it faces as a result of its activities.

As the Staff seems to acknowledge, quantifying risks such as a “slowdown in global economic activity and wider economic contraction in the corporate sector” and “indirect effects resulting from the impact on Western European banking sectors” is very difficult if not impossible.  There is also great uncertainty surrounding such events, including any potential credit event or partial or full breakup of the European Monetary Union.  Citi notes this uncertainty repeatedly, including in its 2011 Form 10-K and 2Q 2012 Form 10-Q:

“…the impact of any such event is highly uncertain and will be based on the specific facts and circumstances. As a result, there can be no assurance that the various steps Citi has taken to protect its businesses, results of operations and financial condition against these events will be sufficient. In addition, there could be negative impacts to Citi’s businesses, results of operations or financial condition that are currently unknown to Citi and thus cannot be mitigated as part of its ongoing contingency planning.”

We will continue to update our disclosures to reflect the material risks associated with Citi’s country risk exposures, as we did in the 2Q 2012 Form 10-Q related to redenomination and devaluation risks.

While we believe we have provided the disclosure requested by the Staff, as discussed above, in its future filings, Citi intends to include disclosure similar to the below as part of the “Overview” section of its “Country Risk” disclosures:

CITI B-8

“As noted in the “Managing Global Risk — Risk Management — Overview” section of the 2011 Form 10-K, Citigroup has instituted a comprehensive risk management process to monitor, evaluate, and manage the principal risks it assumes in conducting its activities, which includes the credit, market and operational risks associated with Citi’s country risk exposures.  The risk management organization is structured to facilitate the management of risk across three dimensions:  businesses, regions and critical products.  The Chief Risk Officer monitors and controls major risk exposures and concentrations across the organization, and subjects those risks to alternative stress scenarios in order to assess the potential economic impact they may have on Citi.  Citi’s independent risk management, working with input from the businesses and finance, provides periodic updates to senior management on significant potential areas of concern across Citi that can arise from risk concentrations, financial market participants, and other systemic issues, such as Eurozone debt issues and other developments in the European Monetary Union.  These areas of focus are intended to be forward-looking assessments of the potential economic impacts to Citi that may arise from these exposures.”

Redenomination and Devaluation Risk, page 95

10.We note your disclosure that you have purchased credit protection primarily from investment grade, global financial institutions predominantly outside of the GIIPS, and that to the extent that purchased credit protection is available in a redenomination/ devaluation event, any redenomination/devaluation exposure could be reduced. Please tell us and expand your disclosure as appropriate to address whether you have reviewed these contracts to determine whether this event is covered. If so, please provide updated disclosure indicating the extent to which your purchased credit protection contracts provide coverage in this event and how the particular provisions operate.

We advise the Staff that we have performed, and will continue to perform, detailed legal reviews of our CDS contracts on sovereign, financial institution and corporate entities to assess whether the CDS contracts would be available for utilization in a redenomination and devaluation event.  Similar analyses are being performed by the International Swaps and Derivatives Association (“ISDA”).  These legal reviews focus on the ISDA Credit Derivatives Definitions as well as additional facts and circumstances that could accompany a redenomination event, such as effective haircuts on the underlying exposures from a significant devaluation of the new local currency, payment or exchange controls, and other factors.

We supplementally advise the Staff that Citi considered including additional disclosure on the availability of purchased CDS contracts as part of its 2Q 2012 Form 10-Q “Redenomination and Devaluation Risk” disclosures.  However, it was determined that such disclosure was not appropriate as it remains highly uncertain (i.e., it is dependent on future fact patterns and ultimately subject to determination by the ISDA Determinations Committee) and could be misconstrued or misinterpreted as more definitive than it is.  More importantly, however, Citi has disclosed its potential redenomination and devaluation exposure assuming no benefit of CDS — meaning any availability of the purchased CDS contracts in a redenomination and devaluation event could only serve to reduce its exposures.  We believe that is the more appropriate disclosure.  Accordingly, Citi currently does not intend to include further disclosure in its future filings regarding the extent to which a redenomination and devaluation event may or may not be covered by its purchased CDS contracts unless there is clarity from a legal and industry perspective.

Notes to Consolidated Financial Statements

Note 14 — Goodwill and Intangible Assets, page 152

11.We note that as of January 1, 2012, you reorganized your reporting units and reassigned goodwill using a relative fair value approach. We also note your response to prior comment 3 from your letter dated March 3, 2009 that you use a risk capital approach as the basis for allocating your total book value to each of your reporting units. Your response indicated that risk capital is the amount

CITI B-9

of capital required to absorb potential unexpected economic losses resulting from extremely severe events over a one-year time period, and also indicated that you allocate every dollar of Citigroup book value to the reporting units as part of the goodwill impairment test and that no net assets are held at the corporate level. Please respond to the following:

·Tell us whether you have made any changes in how you determine the carrying value of your reporting units for purposes of your goodwill impairment testing.

We advise the Staff that there have been no changes made in how we determine the carrying value of the reporting units for purposes of goodwill impairment testing. The carrying value for each reporting unit is derived by allocating reported Citigroup equity using a risk capital approach (as disclosed and defined on p. 67 of the 2011 Form 10-K and as noted by the Staff in its comment above).

·To the extent the total capital allocated to your reporting units is more or less than the total shareholders’ equity of the company, please explain how you account for any difference and quantify such amount. In this regard, clarify whether the carrying value assigned to your reporting units is always equal to the risk capital determined for each reporting unit, or whether any excess or deficit of total shareholders’ equity of the Company over the total calculated risk capital amounts for each of the reporting units is allocated to the reporting units during certain periods, and if so, how.

As we previously informed the Staff, the assigned carrying values of Citigroup’s reporting units, including Corporate/Other (together, the “components”), is equal to Citi’s total shareholders’ equity.  The assigned carrying values of the components are the sum of the goodwill, intangibles and risk capital specific to each component, adjusted such that the total allocated equity is equal to Citigroup’s reported stockholders’ equity on the “as of” date. The overall adjustment needed is calculated as follows: (a) total Citigroup equity, less (b) total Citigroup accounting goodwill and intangibles, less (c) total Citigroup risk capital. The overall adjustment is then allocated pro rata to each component based on the risk capital approach such that the sum of the component carrying values equals Citi’s total shareholders’ equity.

·Expand your disclosure in future filings to discuss how the carrying values of your reporting units are determined. Additionally, to the extent that you do not determine the individual assets and liabilities of your reporting units for purposes of determining the carrying value of the reporting unit, please expand your discussion of how you perform Step 2 of the goodwill impairment test under your methodology when you are required to do so.

In our future Form 10-K filings, we will expand our disclosures related to the determination of reporting unit carrying values. In addition, to the extent that we are required to perform Step 2 of the goodwill impairment test, we will expand our disclosures in future Form 10-K filings related to how Step 2 is performed given our methodology for establishing the carrying values for our reporting units. Specifically, our disclosure will indicate that if Step 2 of the goodwill impairment test is required, the Step 1 carrying value for the respective reporting unit is used as the basis for performing Step 2 of the impairment test, but all actual assets and third-party liabilities of that reporting unit will be identified. If there is a difference between the Step 1 carrying value and the net assets identified, it is assumed that the debt funding should be adjusted (e.g., if the Step 1 carrying value is greater than the net assets identified, it is assumed that intercompany funding is required) to ensure that the Step 2 net assets are equal to the Step 1 carrying value.

Note 16 — Changes in Accumulated Other Comprehensive Income (Loss), page 156

12.We note that ASC 220-10-45-12 requires that the amount of income tax expense or benefit allocated to each component of other comprehensive be shown either on the face of the statement that those items are displayed on in the notes to the financial statements. We also note that in certain places,

CITI B-10

such as pages 181-182, you have disclosed parts of certain of the components in order to calculate the tax effect, but you do not appear to show the tax amount in total or for each of the components. Please revise future filings to include this disclosure, or advise where the disclosures have been provided. To the extent that you believe the disclosures have been provided throughout the document, please revise to present the information in a more summarized way so the amounts can be more easily determined.

Citi confirms for the Staff that, in future filings, we will include pretax and after-tax disclosures within our Changes in Accumulated Comprehensive Income (Loss) Note.

Note 17 — Securitizations and Variable Interest Entities, page 158

Mortgage Securitizations — Citicorp, page 165

13.We note your disclosure of the gains/(losses) recognized on the securitization of U.S. agency-sponsored mortgages for Citicorp on page 165 and for Citi Holdings on page 167. Tell us why the gains recognized upon securitization of U.S. agency-sponsored mortgages by Citicorp were so much lower than the gains recognized by Citi Holdings. In this regard, we note that Citicorp securitized $26.9 billion of U.S. agency-sponsored mortgages during the six months ended June 30, 2012 and recognized $5.9 million in gains, but Citi Holdings securitized $300 million of U.S. agency-sponsored mortgages during the same period, and recognized gains of $30.8 million.

We advise the Staff that all new mortgage originations occur through Citicorp, and we have elected the fair value option for those mortgage loans that are originated for sale. Therefore, the securitizations of U.S. agency-sponsored mortgages by Citicorp are primarily associated with newly originated mortgage loans that are accounted for at fair value, which results in insignificant gains (losses) being recognized upon securitization, relative to the dollar amount of the securitized loans.  The U.S. agency-sponsored mortgage loans securitized by Citi Holdings are not newly originated, and we have not elected the fair value option for them. Accordingly, these loans are classified as held for sale and are accounted for at the lower of cost or market prior to securitization, resulting in the recognition of relatively larger gains upon securitization, as compared with U.S. agency-sponsored mortgage securitizations in Citicorp.

Mortgage Servicing Rights, page 168

14.We note your disclosure that during the second quarter of 2012, the fair value of your mortgage servicing rights (MSRs) was affected by an increase in servicing costs resulting from a change in the mix of loans serviced. Please tell us the effect of this change, and clarify how the mix of loans you serviced changed during the second quarter of 2012.

Consistent with the discussion in the “Mortgage Servicing Rights” section of “North America Consumer Mortgage Lending” under “Managing Global Risk — Credit Risk” on p. 67 of the 2Q 2012 Form 10-Q, as the mix of remaining sold and serviced loans has gradually shifted to more delinquent, non-performing loans, the cost for servicing those loans has increased.  Specifically, lower cost (i.e., better performing) sold and serviced loans are running off (primarily as a result of payoffs) at a higher rate than new origination volume and the run off (primarily as a result of charge-offs) of the higher cost (i.e., delinquent) sold and serviced loans.  The net result is that a higher portion of the sold and serviced loan population consists of higher cost (i.e., delinquent) loans.  We advise the Staff that the effect of this change in loan mix and higher servicing costs resulted in a [***] decrease in the fair value of the MSR in the second quarter of 2012.  An un-redacted copy of this letter was submitted to the Staff pursuant to the Commission’s Rule 83 (17 C.F.R. § 200.83).

[***] The following information has been redacted in accordance with Citigroup’s request for confidential treatment.

Note 19 — Fair Value Measurement, page 185

CITI B-11

15.We note your disclosure on page 186 that you apply market valuation adjustments, including adjustments to account for the size of the net open risk position, consistent with the market participant assumptions and in accordance with the unit of account. We also note your disclosure on page 113 that as a result of the adoption of ASU 2011-04 on January 1, 2012 you released reserves of $125 million related to valuation adjustments where the size of the Company’s position was a characteristic. Please clarify your accounting policy and tell us whether you continue to record any adjustments to account for the size of the Company’s position.

As noted by the Staff and as disclosed, upon adoption of ASU 2011-04 on January 1, 2012, Citi released reserves of approximately $125 million. These reserves primarily related to size-based adjustments on Citi’s cash positions. As a result of the adoption of ASU 2011-04 and consistent with the standard, Citi is no longer permitted to record valuation adjustments for cash instruments and derivatives that are not accounted for under the portfolio exception in paragraph 820-10-35-18D.

Pursuant to paragraph 820-10-50-2D, Citi discloses its accounting policy election to use the portfolio exception in paragraph 820-10-35-18D, which permits a reporting entity to measure the fair value of a group of financial assets and financial liabilities on the basis of the price that would be received to sell or transfer the net open risk position. Paragraph 820-10-35-18D states that “a reporting entity shall measure the fair value of a group of financial assets and financial liabilities consistently with how market participants would price the net risk exposure at the measurement date.” Based on that guidance, when Citi makes the accounting policy election in paragraph 820-10-35-18D, the unit of account is the net open risk position of a portfolio of derivative assets and derivative liabilities.

For most derivatives, Citi makes an accounting policy election, in accordance with Topic 820, to measure the fair value of a portfolio of derivative assets and derivative liabilities on the basis of the net open risk position. As such, pursuant to paragraph 820-10-35-36B, Citi continues to make adjustments to account for the size as a characteristic of its net open risk position, consistent with market participant assumptions and the unit of account.

We believe our existing disclosure on pps. 185 and 186 of the 2Q 2012 Form 10-Q clearly describes the above accounting policy on the application of market valuation adjustments due to the size of Citi’s position and that those adjustments are consistent with the unit of account and market participant assumptions.

Valuation Techniques and Inputs for Level 3 Fair Value Measurements, page 202

16.We note your disclosure on page 203 of the range of high and low values of each significant unobservable input for level 3 assets and liabilities. In addition to the range disclosed, please consider revising your disclosure in future filings to provide a weighted average of the significant unobservable inputs reported, similar to the illustration provided in ASC 820-10-55-103, and state your basis for calculating the weighted average. For example, is the average weighted by the notional or principal amount or by another weighting measure?

We advise the Staff that, when Citi was preparing for the adoption of ASU 2011-04, we considered providing a weighted average of our significant unobservable inputs in the disclosure of Level 3 fair value measurements.  However, we ultimately decided not to do so as we did not find the information to be decision useful to a reader of our financial statements.  For assets that are cash instruments, such as bonds or loans, where price is generally both the most utilized valuation technique and unobservable input, a weighted average of prices of different securities, in isolation, does not provide meaningful information.  For example, the weighted average price of a portfolio of equity securities does not provide meaningful information because the prices of individual securities are independent and are not comparable to one another.  For derivative assets and derivative liabilities as well as those cash assets and liabilities that contain embedded derivatives, such as structured notes or securities purchased/sold under agreements to resell/repurchase, it is even more difficult to provide a meaningful average.  For example, the foreign currency input correlation may

CITI B-12

include the correlations between many currency pairs, such as correlation between two G10 currencies, between a G10 currency and an emerging market currency or between two emerging market currencies.  A weighted average of these currency correlations would not be meaningful and may be misrepresentative of the portfolio characteristics.

We note that the example disclosure in paragraph 820-10-55-103 does not provide a weighted average of inputs for the one derivative type presented, credit contracts.  Additionally, the example does not provide a weighted average of the offered price input in the collateralized debt obligation class.  As a result, we believe that our disclosures are consistent with the example.  To the extent that a weighted average of our significant unobservable inputs is deemed useful for certain classes of instruments in the future, we will consider providing such information.

17.In those situations where multiple valuation techniques are used in the valuation of certain classes of financial assets or liabilities, please consider quantifying the fair values determined under each model. For example, quantify how much of the $12.0 billion of credit derivatives were valued using an internal model as opposed to a price-based approach.

Citi will revise the table on valuation techniques and inputs for Level 3 fair value measurements in future filings to quantify the fair values determined under each valuation technique used.

18.We note that you use a price-based methodology for certain of your classes of financial assets and liabilities, including mortgage-backed securities, state and municipal, foreign government, corporate and other debt securities, asset-backed securities interest rate contracts, credit derivatives, loans, and shorter-term borrowings and long-term debt. We also note that in each of these cases, “price” is listed as at least one of the significant unobservable inputs. Please clarify how the price-based methodology works, and why the “price” is a significant unobservable input. Additionally, please tell us why, in nearly every case, $0.00 is the low end of the range input for the price assumption and the high is typically in the low to mid $100.00 range.

We advise the Staff that, in certain cases, the individual business unit does not use a model to value an instrument classified in the Level 3 fair value category, instead, a price-based methodology is applied. Price-based methodologies are valuation techniques that utilize prices or other market information obtained from transactions. Where such “price” inputs are considered unobservable, the fair value measurements are classified as Level 3. Unobservable “price” inputs may include the last trade of the same security, prices of similar securities, or external broker quotes. There are various reasons why a “price” input would be deemed unobservable, including where the last trade used to value an instrument is stale, a significant adjustment to the price of a similar security is necessary due to differences in the terms of the actual security being valued, or there is insufficient market data (e.g., few immaterial recent transactions) to corroborate the valuation.  In these and other similar cases, a price will be deemed an unobservable input and the valuation technique will be designated as “price-based.”

The minimum value for the price input can be zero if the class of financial assets includes a security, such as a subprime CDO, that is not receiving any principal or interest and currently is written down to zero. The high end of the range of the price input reflects the structured nature of certain securities. For example, the price of $159.63 was for a medium-term bond linked to a derivative payoff, resulting in a large price premium.

Sensitivity to Unobservable Inputs and Interrelationships between Unobservable Inputs, page 204

19.We note that your sensitivity of unobservable inputs is somewhat generic and not linked specifically to the level 3 instruments which contain these inputs, and that you do not discuss how the specific instruments could directionally be impacted by changes in the specific unobservable inputs. For example, your disclosure indicates that changes in correlation levels can have a major impact, favorable or unfavorable, on the value of the instrument. Please explain in more detail the factors

CITI B-13

that would drive a favorable or unfavorable impact, and the types of instruments that would most be affected by those factors. Similarly, we note that additional disclosures could be added to the volatility section to indicate the types of instruments (or factors that exist in the instruments) that would benefit from an increase in volatility versus a decrease in volatility. We note similar comments apply to each discussion of the sensitivity of the unobservable inputs. Please refer to the guidance in ASC 820-10-55-106 for an illustration of how the disclosure could be expanded to be linked more directly to your instruments, and expand your disclosures in future filings to address these points.

In response to the Staff’s comment, in future filings, Citi intends to revise its sensitivities disclosures to provide more specific disclosure as to the factors that would drive a favorable or unfavorable impact and the types of Level 3 instruments that would be most affected by such factors.  Citi intends to include disclosure on volatility similar to the below in future filings. Citi intends to provide similar disclosures, as applicable, regarding other disclosed sensitivities.

Volatility

Volatility represents the speed and severity of market price changes and is a key factor in pricing options. Typically, these instruments can become more expensive if volatility increases. For example, as an index becomes more volatile, the cost to Citi of maintaining a given level of exposure increases because more frequent rebalancing of the portfolio is required. Volatility generally depends on the tenor of the underlying instrument and the strike price or level defined in the contract. The general relationship between changes in the value of a portfolio to changes in volatility also depends on changes in interest rates and the level of the underlying index. Generally, long option positions (assets) benefit from increases in volatility, whereas short option positions (liabilities) will suffer losses. In addition, some instruments are more sensitive to changes in volatility than others. For example, an at the money option would experience a larger percentage change in its fair value than a deep in the money option. In addition, the fair value of an option with more than one underlying security (for example, an option on a basket of bonds) depends on the volatility of the individual underlying securities as well as their correlations.

Items Measured at Fair Value on a Nonrecurring Basis, page 205

20.We note that you have classified impaired loans as level 2 in the fair value hierarchy, and have disclosed that the fair value is determined based on quoted prices for similar assets, adjusted for the attributes of the loan, or based on the fair value of the collateral, which is typically estimated based on the quoted market prices if available, appraisals or other internal valuation techniques. Please tell us in more detail how you determined that the techniques used for these impaired loans qualified as level 2 in the fair value hierarchy. For example, describe the types of impaired loans and the market information used in the analysis to support a level 2 classification.

As the Staff is aware, under the ASC 820 fair value hierarchy, level 2 includes instruments with quoted prices for similar assets, where all significant inputs are observable, while in level 3, at least one significant input is unobservable.  We advise the Staff that the disclosure for “Items Measured at Fair Value on a Nonrecurring Basis” in the 2Q 2012 Form 10-Q includes loans held for investment whose carrying amount is based on the fair value of the underlying collateral, which is primarily real-estate secured loans.  The valuation techniques for these loans utilize appraisals of the underlying property that are based on sales of similar properties in the geographic area.  For those valuations where no significant unobservable input is used to adjust the appraised value, the loan is classified as level 2.  For those valuations where a significant unobservable input is utilized to adjust the appraised value, the loan is classified as level 3.

21.Additionally, we note ASC 820-10-50-2(bbb) requires disclosure of the inputs used in the fair value measurement for both level 2 and level 3 nonrecurring fair value measurements, and we were unable to locate this information for any of your nonrecurring fair value measurements. Please advise, or revise to provide this information in future filings.

CITI B-14

We advise the Staff that on p. 205 of our 2Q 2012 Form 10-Q, we disclosed the following:

“The fair value of loans-held-for-sale is determined where possible using quoted secondary-market prices.  If no such quoted price exists, the fair value of a loan is determined using quoted prices for a similar asset or assets, adjusted for the specific attributes of that loan.  Fair value for the other real estate owned is based on appraisals.  For loans whose carrying amount is based on the fair value of the underlying collateral, the fair values depend on the type of collateral.  Fair value of the collateral is typically estimated based on quoted market prices if available, appraisals or other internal valuation techniques.”

In future filings, we will expand our disclosures of the valuation techniques and inputs used for our significant level 2 and level 3 nonrecurring fair value measurements, as requested by the Staff.

Note 22 — Contingencies, page 219

Interchange Litigation, page 220

22.We note your disclosure regarding the terms of the class settlement contemplated by the MOU, including a rebate to merchants participating in the class settlement of 10 basis points on interchange collected for a period of eight months by Visa and MasterCard networks. Please tell us how you plan to account for this aspect of the contemplated settlement. Additionally, please tell us whether this represents the last aspect of the covered litigation such that the Visa escrow account will be terminated. If so, please tell us how this will affect any shares you continue to hold in Visa going forward.

With respect to the accounting for the rebate provision of the class settlement, we advise the Staff that Citi determined that the estimated cost of the rebate should be accrued during the second quarter of 2012 because it was probable and estimable.  As a result, Citi recorded a [***] litigation reserve increase impacting expenses in the second quarter of 2012, which is Citi’s estimate of the rebate amount to be paid by Citi and which predominantly impacted Citi’s branded cards business in North America.

We further advise the Staff that the timing of the “termination” of the Visa escrow account is to be determined by Visa.  We believe there is likely to be further litigation with merchants that opt out of the class settlement, as some major retailers have stated publicly that they are not satisfied with the settlement, which suggests that they may opt out and pursue their own litigation.  It is our understanding that the escrow account is intended to cover that litigation.  We also believe that there is at least one other small remaining case that implicates the Visa escrow account.  As a result, we do not believe that the class settlement contemplated by the MOU represents the last piece of litigation covered by the Visa escrow account. However, upon termination of the Visa escrow account, we believe the Visa B shares held by Citi would become eligible for accounting under FAS 115 as codified under ASC 320, resulting in recognizing the Visa B shares at fair value with changes in fair value recognized in either other comprehensive income or net income, depending on whether they will be classified as available-for-sale or trading securities.

[***] The following information has been redacted in accordance with Citigroup’s request for confidential treatment.

Interbank Offered Rates—Related Litigation and Other Matters, page 220

23.We note your brief discussion of the ongoing LIBOR investigations and lawsuits on page 220. In the future, rather than solely referring to docket numbers, please revise your disclosure to provide more detail regarding each outstanding matter. Also, please confirm that you will provide more information regarding new developments as they relate to the company and its affiliates.

In future filings, Citi will revise its disclosures on LIBOR matters to provide additional detail regarding each matter disclosed and confirms that it will update these disclosures for material developments, as applicable and appropriate.

CITI B-15

*********

In connection with responding to the Staff’s comments, Citigroup acknowledges that it is responsible for the adequacy and accuracy of the disclosures in its filings; that the Staff’s comments or changes to disclosure in response to Staff comments do not foreclose the SEC from taking any action with respect to the filing; and that Citigroup may not assert Staff comments as a defense in any proceeding initiated by the SEC or any person under the federal securities laws of the United States.

Sincerely,

/s/ John C. Gerspach

John C. Gerspach

Chief Financial Officer

CITI B-16