SecProbe.io

CORRESP
1
filename1.htm

August 20, 2007

Kevin W. Vaughn

Branch Chief

Securities and Exchange Commission

120 F Fifth Street, N.E.

Washington, DC 20549

  RE:

  Citigroup Inc.

  Form 10-K for Fiscal Year Ended December 31,
  2006

  Filed February 23, 2007

  File No. 001-09924

Dear Mr. Vaughn:

Enhancement of the
overall disclosures in filings by Citigroup Inc. (“Citigroup,” the “Company,”
or “we”) is an objective that we share with the Staff and one that we consider
in all our filings.  This letter sets
forth the responses of Citigroup to the comments of the staff of the SEC
contained in your letter dated July 3, 2007.

Events in 2005

Bank and Credit Card
Customer Rewards Costs, page 11

1.               If material, please
revise the footnotes to your financial statements in your future filings to
disclose your policies for these programs.
Specifically address how such amounts are reported on the face of your
financial statements for each of your on-balance sheet portfolio and your off-balance
sheet portfolio.  If you determined such
amounts are not material for disclosure, tell us the amounts and provide us
with your policies and presentation in your response.

Citigroup’s
Response: We consider the credit card customer rewards costs
to be significant and have added the following language in Note 1 to our
consolidated financial statements, included in our 2007 second quarter Form
10-Q:

“The
Company recognizes all reward costs for bank and credit card customers as
incurred as a reduction of Commissions and fees and Other revenue.”

Selected Revenue and
Expense Items

U.S. Consumer Lending,
page 26

2.               Please tell us and
disclose in future filings the degree of your exposure to sub prime mortgages
in each your mortgage loans held for investment and through
securitizations.  Specifically address
the credit quality trends within this product type and the impact of these
trends on your loan loss provision.

Citigroup’s Response: Our U.S. Consumer business does not
refer to a definition of “subprime,” but rather uses the terms “non-prime” and “prime”
to categorize our loans. We have included the following language in the
Management’s Discussion and Analysis section of our second quarter 2007 Form
10-Q:

“U.S. Consumer Mortgage Portfolio

The Company’s U.S. Consumer Mortgage portfolio
consists of both first and second mortgages.
As of June 30, 2007, approximately 85% of the Company’s first mortgage
portfolio had a FICO (Fair Isaac Corporation) credit score of at least 620 at
origination.  Approximately 72% of the
first mortgage portfolio had a loan-to-value (LTV) ratio of 80% or less at
origination.

In the Company’s second mortgage portfolio, there
were substantially less than 1% of loans at June 30, 2007 with a FICO score of
less than 620 at origination. Approximately 48% of the second mortgage
portfolio had an LTV ratio of 80% or less at origination.

In light of increased delinquencies, the Company
has increased reserves for loans in its second mortgage portfolio during the
first six months of 2007.  There were
minimal changes in the composition of the U.S. Consumer Mortgage portfolio from
March 31, 2007 to June 30, 2007.”

We did not disclose the amount of mortgage-backed
securities and residual interests collateralized by non-prime mortgages held by
U.S. Consumer due to immateriality.

Consolidated Financial
Statements

Note 1.  Summary of Significant Accounting Policies

Loans Held for Sale, page
111

 2

3.               In future filings
beginning in your next Form 10-Q, please revise your disclosures to include the
following information:

·                  You disclose
that you classify a loan as held for sale when it is identified for sale.  Your correspondence filed with the Commission
on June 6, 2006 indicates that you determine the classification of loans upon
origination or purchase.  Please revise
your disclosures to clarify what your policies are for determining the
classification of loans and credit card receivables on your balance sheet upon
origination or purchase.

·                  Please expand
your disclosures to include the key assumptions management uses in determining
the classification of loans and credit card receivables upon origination or
purchase and specifically disclose what you consider to be the foreseeable
future under paragraph 8(a) of SOP 01-6 for each category of loans and
receivables.

·                  Disclose the
impact of these policies on the operating and investing section of your
statement of cash flows.

Citigroup’s Response: As we agreed in our phone conversation with the SEC
staff on June 14, 2007, the following disclosure has been added to Citigroup’s
Form 10-Q for the period ended June 30, 2007, as noted:

“Loans are classified upon origination or acquisition as
either held for investment or held for sale. This classification is based on
management’s intent and ability with regard to those loans.

Substantially all of the consumer loans sold or securitized
by Citigroup are U.S. prime mortgage loans or U.S. credit card
receivables.  The practice of the U.S.
prime mortgage business has been to sell all of its loans except for
nonconforming adjustable rate loans. U.S. prime mortgage conforming loans are
classified as held for sale at the time of origination. The related cash flows
are classified in the Consolidated Statement of Cash Flows in the cash flows
from operating activities category on the line “Change in loans held for sale.”

U.S credit card receivables are classified at origination as
loans held for sale to the extent that management does not have the intent to
hold the receivables for the foreseeable future or until maturity. The U.S.
credit card securitization forecast for the three months following the latest
balance sheet date is the basis for the amount of such loans classified as held
for sale. Cash flows related to U.S. credit card loans classified as held for
sale at origination or acquisition are reported in the cash flows from
operating activities category on the line “Change in loans held for sale.”

 3

Loans that are held for investment
are classified as Loans, net of unearned income on the Consolidated Balance
Sheet, and the related cash flows are included in the cash flows from investing activities
category in the Consolidated Statement of Cash Flows on the line “Changes in
loans.”  However, when the initial intent
for holding a loan has changed from held for investment to held for sale, the
loan is reclassified to held for sale, but the related cash flows continue to
be reported in cash
flows from investing activities in the Consolidated Statement of Cash Flows on the
line “Proceeds from sales and securitizations of loans.”

Note 4.  Business Segments

4.               In your Five Year
Summary of Selected Financial Data, you present measures of Return on Risk
Capital and Return on Invested Capital.
In your correspondence filed with the commission on June 6, 2006, you
represented that these measures are used, in part, to assess long-term growth
of your segment businesses.  Further, you
stated in that correspondence that you would add disclosures regarding these
measures to your segment footnotes.  We
were unable to locate the proposed disclosures referred to in your
response.  Please tell us where such
disclosures were made in your footnotes or revise your future filings beginning
with your next Form 10-Q to add those disclosures.

Citigroup’s Response: The
disclosures we referred to in our June 6, 2006 response to the SEC were made in
Citigroup’s 2006 Form 10-K, where we footnoted each disclosure of Risk Capital
and Invested Capital in each of our segment and product disclosures to sub-note
(6) on the Five-Year Summary (page 3 in our Form 10-K) in the Management’s
Discussion and Analysis section. Also, we expanded our discussion of Risk
Capital and related returns in our Five Year Summary and in our Managing Global
Risk section on page 59 of our Form 10-K.

Note 23, Derivative
Activities, page 148

5.               We note from your
disclosures that you utilize regression analysis for multiple types of fair
value and cash flow hedges.

·                  For each type of
asset and liability and each specific risk being hedged, please tell us the
regression outputs you consider when assessing whether these hedges are expected
to be highly effective during the period that the hedge is designated.

 4

·                  Tell us the
extent to which you exclude any outputs from your consideration of
effectiveness.

·                  If you do
exclude certain outputs, tell us your basis for doing so.  Specifically reference the appropriate
sections of SFAS 133, including any applicable DIG Issues that support this
approach.

Citigroup’s Response: As of
December 31, 2006 and March 31, 2007, in hedges that utilize regression
analysis to test effectiveness, Citigroup uses the following regression outputs
to ensure that the relevant hedging relationship has either been effective
retrospectively or is expected to remain effective prospectively:

  Hedge Type

  Risk being hedged

  Regression outputs considered

  Assets

  Floating rate available-for-sale debt securities

  Cash Flow

  Benchmark interest rates

  r2, slope, T-Test

  Floating rate available-for-sale debt securities

  Cash Flow

  Overall cash flows

  r2, r

  Fixed rate available-for-sale debt securities

  Fair Value

  Benchmark interest rates

  r2, slope, T-Test

  Held for sale mortgage loans

  Fair Value

  Overall fair value

  r

  Liabilities

  Forecasted issuances of short term fixed rate
  liabilities subject to rollover

  Cash Flow

  Benchmark interest rates

  r2, slope, T-Test

  Federal funds purchased

  Cash Flow

  Overall cash flows

  r2, slope, T-Test

With the exception of the hedge of held for sale
mortgage loans, the coefficient of determination (r2) is the primary
consideration when determining whether the hedging relationship is expected to
be highly effective and/or has been highly effective. As noted above, depending
on the hedging relationship, we also consider the slope coefficient and the
T-test. Citigroup does not specifically exclude any regression outputs from
consideration of effectiveness.

We have decided to discontinue hedge accounting
under SFAS 133 and elect the Fair Value Option under SFAS 159 for held for sale
mortgage loans in the third quarter of 2007.

In general, our SFAS 133 hedging relationships do
not involve significant basis risk, because the hedged items are of a “plain
vanilla” nature and the risks being

 5

hedged are widely quoted in highly liquid markets.
Thus, we are able to construct SFAS 133 hedging relationships where the
primary terms of the hedged item and hedging instrument are closely aligned.
For that reason, we believe that it is reasonable to expect that any regression
performed between the two sets of data will be statistically valid.

6.               Based on your
disclosures in Note 23, you employ several different approaches in assessing
effectiveness for your hedges, including regression analysis, the hypothetical
derivative method, the dollar offset ratio method, the short-cut method, and
the matching critical terms method.  To
the extent that you used other methodologies to assess hedge effectiveness,
made qualitative assumptions of effectiveness without performing quantitative
tests, used multiple types of tests on same hedge relationship to determine a
test is effective even though one or more tests may have failed, or other
statistical methods, please address the following:

·                  For each type of
long-haul relationship (fair value, cash flow, net investment hedge, etc.)
entered into during the periods presented, please describe for us the quantitative
and qualitative measures you use to assess effectiveness both at inception and
on an ongoing basis.

·                  Specifically
reference the appropriate sections of SFAS 133, including any applicable DIG
Issues, which support the use of each test.

Citigroup’s
Response: As of December 31, 2006 or March 31, 2007, we did
not use any methodologies to assess hedge effectiveness other than those listed
above and disclosed in Note 23 of our 2006 Annual Report. We did not make
qualitative assumptions of effectiveness without performing quantitative tests.
We did not use multiple types of tests (for example, both dollar offset ratio
and regression analysis) on the same hedging relationship to determine that a
hedge is effective if one test may have failed. We did not use statistical
methods other than regression.

7.               To help us gain a
better understanding of your hedge accounting process, please provide us with
the following additional information:

·                  Describe the
process by which derivative instruments are obtained by the various segments
for hedge accounting purposes; for example, tell us which group or groups (i.e.
Treasury, Trading Desk etc.) are responsible for obtaining these derivative
instruments and describe the process by which they are obtained;

 6

·                  Tell us if the responsible
groups obtain the derivative instruments from internal or external sources or
both;

·                  Tell us whether
you have designated an internal derivative as the hedging instrument in any
hedging relationship, and if so, describe the nature of the hedging
relationship, methodology used to assess effectiveness, and the applicable
guidance relied upon to support the approach; and

·                  Tell us how the
various segments determine that the conditions for hedge accounting are
achieved.

Citigroup’s Response: The primary
segments that employ SFAS 133 hedge accounting are Global Consumer, Markets and
Banking, and Corporate/Other. Each hedging relationship requires an external
derivative to be designated as the hedging instrument. We do not designate
internal derivatives as the hedging instrument on a consolidated basis,
including in the limited permitted circumstances described in paragraphs 40A
and 40B of SFAS 133.

Each segment (hedging unit) is responsible for
determining whether it will enter into a derivative instrument directly with an
external counterparty or through a trading desk within Markets and Banking,
which then executes a derivative instrument with an external source. For
example, if a Citigroup hedging unit decides to execute derivative instruments
through the trading desk in Markets and Banking, the trading desk enters into a
derivative instrument with an external source and that third-party derivative
instrument serves as the hedging instrument under SFAS 133.

Each hedging unit is responsible for determining
that the conditions for hedge accounting are achieved. Citigroup achieves this
through a formal process requiring the completion, approval, implementation and
review of “hedge programs.” No hedge accounting activity is permitted without
such prior approvals. Hedge accounting documentation is prepared for each
hedging relationship under each program and Audit and Risk Review (internal
audit) also reviews operational compliance with the hedge programs. Segment and
Corporate Accounting Policy approves, monitors, and periodically reviews hedge
programs to ensure desired hedging strategies continue to be met.

8.               In your
correspondence filed on September 6, 2006, you indicated that you utilize the
short-cut method of assessing hedge effectiveness for certain of your cash flow
hedges.  You disclosure on page 149 does
not address the use of short-cut for any of your cash flow hedges, but does
refer to your efforts made to match all critical terms of the hedged item and
the hedging derivative at inception and on an

 7

ongoing basis to eliminate hedge ineffectiveness
related to certain of your cash flow hedges.

·                  In future
filings, please revise to clarify in your disclosures if you are using the