Flagstar Bancorp (FBC) CORRESP Correspondence with SEC

December 30, 2009

Mr. Christian Windsor
Special Counsel
Division of Corporation Finance
Securities and Exchange Commission
100 F Street, N.E.
Washington, D.C. 20549
Mail Stop 4720

Re:   Flagstar Bancorp, Inc.
Form 10-K for the fiscal year ended December 31, 2008
Form 10-Q for the quarterly period ended March 31, 2009
Form 10-Q for the quarterly period ended June 30, 2009
File No. 001-16577
Form S-3 filed November 3, 2009
File No. 333-162823

Dear Mr. Windsor:

     We are writing with respect to your letter, dated December 17, 2009 (the “Comment Letter”), regarding the Form 10-K for the fiscal year ended December 31, 2008 filed by Flagstar Bancorp, Inc. (the “Company” or “Flagstar”) on March 13, 2009, the Form 10-Q for the quarterly period ended March 31, 2009 filed by Flagstar on May 7, 2009, the Form 10-Q for the quarterly period ended June 30, 2009 filed by Flagstar on August 5, 2009, the Form 8-K filed by Flagstar on December 14, 2009, and the Form S-3 filed by Flagstar on November 3, 2009.

     Our responses to the comments included in the Comment Letter are set forth below, in each case, except Comments 3 through 6, preceded by a reproduction of the corresponding comment.

Form 10-K for the period ended December 31, 2008

Business, page 3

Comment No. 1:	Please refer to your response to comment 1 of our letter dated September 23, 2009 and address the following:
Comment No. 1(a):	Please tell us and revise future filings to provide a tabular presentation of the future reset dates for your ARM loans. It should be clear to a reader how much of these loans will reset in each future period.
Response No. 1(a):	In response to the Staff’s comment, set forth below as of September 30, 2009 is a tabular presentation of the future reset dates

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 2

for our ARM loans. We will include this presentation in our future filings.

“Set forth below as of September 30, 2009 [to be updated for December 31, 2009], are the amounts of the ARM loans in our held-for-investment loan portfolio with interest rate reset dates in the periods noted:

($ in thousands)	1st Quarter		2nd Quarter		3rd Quarter		4th Quarter
2010	$	821,281	$	1,040,208	$	1,084,376	$	1,133,040
2011	$	1,089,338	$	1,188,018	$	1,143,763	$	1,229,298
2012	$	1,220,741	$	1,455,642	$	1,467,275	$	1,449,314
Later years (1)	$	1,444,6441	$	1,487,400	$	1,515,521	$	1,505,686

(1) Later years reflect one reset period per loan.

Comment No. 1(b):	Please tell us and revise future filings to explain why you believe presenting your underwriting polices and procedures for mortgage and HELOC loans originated prior to 2009 in a separate section from loans originated in 2009 is appropriate, clarifying why you have presented some periods under Banking Operations and some under Home Lending Operations. Alternatively, please revise to include this in one section.
Response No. 1(b):	We have revised, and set forth below, our summary of underwriting policies and procedures for mortgage and HELOC loans for 2009 and prior years in the same section, and have also combined those sections previously set forth separately in Banking Operations and Home Lending Operations. We are also including the additional disclosures noted in the comments above and below.
	Replace the current disclosure under Item 1. BUSINESS — Operating Segments — Home Lending Operation — Underwriting, with the following disclosure:
	In past years, we originated a wide variety of residential mortgage loans, both for sale and for our own portfolio, including fixed rate first and second lien mortgage loans, ARMs, interest only mortgage loans both ARM and fixed, potential negative amortization payment option ARMs (option power ARMs), subprime loans, and HELOCs. We also originated commercial real estate loans for our own portfolio.
	As a result of our increasing concerns about national economic conditions, in 2007 we began to reduce the number and types of loans that we originated for our own portfolio rather than for sale into the secondary market. In 2008, we halted originations of virtually all types of loans for

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 3

our held-for-investment portfolio and focused on the origination of residential mortgage loans for sale. In addition, we increased our monitoring of real estate valuations throughout our held-for-investment portfolio. However, we generally do not obtain updated appraisals or broker price opinions on our mortgage or HELOC loans until such loans enter the foreclosure process.

During 2009, we primarily originated residential mortgage loans for sale that conformed to the respective underwriting guidelines established by each Agency program.

     First Mortgage Loans.

At [September 30, 2009] [to be updated at December 31, 2009], most of our held-for-investment mortgage loans were originated in prior years with underwriting criteria that varied by product and with the standards in place at the time of origination.

Set forth below is a table describing the characteristics of the first mortgage loans at [September 30, 2009] [to be updated at December 31, 2009] in our held-for-investment portfolio, by year of origination (also referred to as the vintage year, or vintage).

	Prior to
($ in thousands)	2007			2007			2008			2009			Total
Unpaid Principal Balance	$	3,311,949		$	1,818,474		$	109,863		$	25,042		$	5,265,328
Average Note Rate		5.35	%		6.26	%		6.29	%		5.64	%		5.69	%
Average Original FICO Score		716			720			692			692			717
Average Original Loan-to-Value Ratio		73.99			74.56			82.97			81.87			74.41
Average Original Combined Loan-to-Value Ratio		77.59			77.51			84.43			87.98			77.76
Underwritten with low or stated income documentation		32	%		56	%		19	%		7	%		40	%

First mortgage loans are underwritten on a loan-by-loan basis rather than on a pool basis. Generally, mortgage loans produced through our production channels are reviewed by one of our in-house loan underwriters or by a contract underwriter employed by a mortgage insurance company. However, certain of our correspondents have delegated underwriting authority but this never comprised more than 15% of the loans. In all cases, loans must be underwritten to Flagstar’s standards. Any loan not underwritten by a Flagstar-employed underwriter must be warranted by the underwriter’s employer, whether it is a mortgage insurance company or correspondent mortgage company.

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 4

Our criteria for underwriting generally included, but were not limited to, full documentation of borrower income and other relevant financial information, fully indexed rate consideration for variable loans, and Agency eligible loan-to-value ratios with full appraisals when required. Variances from any of these standards were permitted only to the extent allowable under the specific Agency’s program requirements. These included the ability to originate loans with less than full documentation and variable rate loans with an initial interest rate less than the fully indexed rate. Mortgage loans were collateralized by a first or second mortgage on a one-to-four family residential property.

In general, loan balances under $1,000,000 required a valid Agency automated underwriting system (“AUS”) response for approval consideration. Documentation and ratio guidelines were driven by the AUS response. A FICO credit score for the borrower was required and a full appraisal of the underlying property that would serve as collateral was obtained.

For loan balances over $1,000,000, traditional manual underwriting documentation and ratio requirements were required as were two years plus year to date of income documentation and two months of bank statements. Income documentation based solely on a borrower’s statement was an available underwriting option for each loan category. Even so, in these cases employment of the borrower was verified under the vast majority of loan programs, and income levels were usually checked against third party sources to confirm validity.

We believe that our underwriting process, which relies on the electronic submission of data and images and is based on an award-winning imaging workflow process, allows for underwriting at a higher level of accuracy and with more timeliness than exists with processes which rely on paper submissions. We also provide our underwriters with integrated quality control tools, such as automated valuation models (“AVMs”), multiple fraud detection engines and the ability to electronically submit IRS Form 4506s, to ensure underwriters have the information that they need to make informed decisions. The process begins with the submission of an electronic application and an initial determination of eligibility. The application and required documents are then faxed or uploaded to our corporate underwriting department and all documents are identified by optical character recognition or our underwriting staff. The underwriter is responsible for checking the data integrity and reviewing credit. The file is then reviewed in accordance with the applicable guidelines established by us for the particular product. Quality control checks are performed by the

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 5

underwriting department using the tools outlined above, as necessary, and a decision is then made and communicated to the prospective borrower.

The following table identifies, at [September 30, 2009] [to be updated at December 31, 2009], our held-for-investment mortgages by major category and describes the current portfolio with unpaid principal balance, average note rate, average original FICO score, average original combined loan-to-value ratio (“CLTV”), the weighted average maturity and the related housing price index. The housing price index (“HPI”) loan-to-value (“LTV”) is updated based on Metropolitan Statistical Area (“MSA”)-level OFHEO data. The only loans categorized as subprime, comprising only 0.1% of the portfolio of first liens, were originated using loan underwriting guidelines that were established by mortgage companies such as Countrywide and New Century. Within the first lien residential mortgage loan portfolio, high LTV loan originations, defined as loans with a 95% LTV or greater, comprised only 6% of our held-for-investment loan portfolio. Our risk of loss on these loans is mitigated because private mortgage insurance was required on the vast majority of loans with LTVs exceeding 80% at the time of origination.

								Average
						Average		Original
	Unpaid		Average			Original		Combined		Weighted		Housing
	Principal		Note			FICO		Loan-to-		Average		Price
($ in thousands)	Balance		Rate			Score		Value Ratio		Maturity		Index
First mortgage loans:
Amortizing:
3/1 ARM	$	285,275		5.10	%		689		81.77		286		78.37
5/1 ARM	$	749,584		5.19	%		717		74.70		301		68.83
7/1 ARM	$	99,481		5.56	%		729		73.60		304		74.59
Other amortizing	$	1,077,412		6.15	%		708		76.30		286		80.89
Interest only:
3/1 ARM	$	404,560		5.40	%		723		81.46		274		82.92
5/1 ARM	$	1,660,441		5.55	%		722		78.79		306		82.83
7/1 ARM	$	137,756		6.13	%		728		74.98		313		90.60
Other interest only	$	561,362		6.15	%		723		78.55		319		93.75
Option ARMs	$	283,334		5.97	%		720		76.85		330		99.36
Subprime	$	6,123		7.54	%		610		87.95		288		100.99
Total first mortgage loans	$	5,265,328		5.69	%		717		77.76		300		82.33
Second mortgages	$	235,096		8.44	%		733		89.28		157		22.08
HELOC’s	$	316,498		5.35	%		739		80.86		77		25.09

In addition to the above information, 39% of the loans underwritten in prior years had income or asset documentation requirements that were less than those now being imposed. Substantially all of those loans were underwritten with verification of employment but with the related job income or personal assets, or both, stated by the borrower without

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 6

verification of actual amount. Those loans may have additional elements of risk since information provided by the borrower in connection with the loan was limited.

The following table sets forth characteristics of those loans in our held for investment portfolio as of [September 30, 2009] [to be updated at December 31, 2009] that were originated with less documentation than is currently required. Loans as to which underwriting information was accepted from a borrower without validating that particular item of information is referred to as “low doc” or “stated,” and loans as to which underwriting information was supported by third party documentation or procedures is referred to as “full doc” and the information therein is referred to as “verified.” Also set forth are different types of loans that may have a higher risk of non-collection than other loans.

	Low Doc
	% of Held-for-			Unpaid
	Investment			Principal
($ in thousands)	Portfolio			Balance
Characteristics
SISA (stated income, stated asset)		3	%	$	214,363
SIVA (stated income, verified assets)		17	%	$	1,400,655
High LTV (i.e., at or above 95%)		0	%	$	22,566
Second lien products (HELOCs, 2nd mortgages)		2	%	$	163,126
Loans with initial teaser rate that have not reset		0	%	$	—
Loan types:
Option ARM loans		2	%	$	199,865
Interest-only loans		15	%	$	1,203,929
Subprime		0	%	$	3,212

     ARMs

     Of the ARM loans held for investment, 94% do not have the potential for negative amortization. These loans were originated using Fannie Mae and Freddie Mac guidelines as a base framework, and the debt-to-income ratio guidelines and documentation typically followed the AUS response. The maximum ratios allowable for purposes of both the LTV ratio and the CLTV ratio, which includes second mortgages on the same collateral was 100%, but subordinate (i.e., second mortgage) financing was not allowed over a 95% LTV ratio. At a 100% LTV ratio, the minimum acceptable FICO score, or the “floor,” was 700, and at lower LTV ratio levels, the FICO floor was 620. All occupancy and specific-purpose loan types were allowed at lower LTVs. At times ARMs were underwritten at an initial rate, also known as the “start rate”, that was lower that the fully indexed rate but only for loans with lower LTV ratios and higher FICO scores. Other ARMs were either underwritten at the note

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 7

rate if the initial fixed term was two years or greater, or at the note rate plus two percentage points if the initial fixed rate term was six months to one year.

Adjustable rate loans were not consistently underwritten to the fully indexed rate until the Interagency Guidance on Nontraditional Mortgage Products, issued by the federal banking regulatory agencies, was released in 2006. Teaser rates, i.e., the initial rate on the loan was discounted from the otherwise applicable fully indexed rate were only offered for the first three months of the loan term, and then only on a portion of Adjustable Rate Mortgages (“ARMs”) that had been the negative amortization payment option available and HELOCs. Due to the seasoning of our portfolio all borrowers have adjusted out of their teaser rates at this time.

Option power ARMs, comprising 6% of the first mortgage portfolio, are adjustable rate mortgage loans that permit a borrower to select one of three monthly payment options when the loan is first originated: (i) a principal and interest payment that would fully repay the loan over its stated term, (ii) an interest-only payment that would require the borrower to pay only the interest due each month but would have a period (usually 10 years) after which the entire amount of the loan would need to be repaid (i.e., a balloon payment) or refinanced, and (iii) a minimum payment amount selected by the borrower and which might exclude principal and some interest, with the unpaid interest added to the balance of the loan (i.e., a process known as “negative amortization”).

Option power ARMS were originated with maximum LTV and CLTV ratios of 95%; however, subordinate financing was only allowed for LTVs of 80% or less. At higher LTV/CLTV ratios, the FICO floor was 680, and at lower LTV levels the FICO floor was 620. All occupancy and purpose types were allowed at lower LTVs. The negative amortization cap, i.e., the sum of a loan’s initial principal balance plus any deferred interest payments, divided by the original principal balance of the loan, was generally 115%, except that the cap in New York was 110%. In addition, for the first five years, when the new monthly payment due is calculated every twelve months, the monthly payment amount could not increase more than 7.5% from year to year. By 2007, option power ARMs were underwritten at the fully indexed rate rather than at a start rate. At [September 30, 2009], we had $283.3 million of option power ARM loans in our held-for-investment loan portfolio, and the amount of negative amortization reflected in the loan balances at September 30, 2009 was $16.0 million.

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 8

Set forth below is a table describing the characteristics of our ARM loans at [September 30, 2009] [to be updated at December 31, 2009] in our held-for-investment portfolio, by year of origination.

	Prior to
($ in thousands)	2007			2007			2008			2009			Total
Unpaid Principal Balance	$	2,864,350		$	899,017		$	50,977		$	14,489		$	3,828,834
Average Note Rate		5.23	%		6.18	%		6.05	%		5.57	%		5.47	%
Average Original FICO Score		717			720			716			683			717
Average Original Loan-to-Value Ratio		74.31			74.95			80.69			77.11			74.55
Average Original Combined Loan-to-Value Ratio		78.24			78.28			84.24			87.31			78.36
Underwritten with low or stated income documentation		34	%		59	%		22	%		12	%		40	%
Borrowers electing to make less than full interest payments		6	%		35	%		4	%		3	%		13	%

��

Set forth below is a table describing [as of September 30, 2009] [to be updated at December 31, 2009] characteristics specifically of option power ARMs, by year of origination:

	Prior to
($ in thousands)	2007			2007			2008		2009		Total
Unpaid Principal Balance	$	82,727		$	200,608		$	—	$	—	$	283,334
Average Note Rate		6.24	%		5.31	%		—		—		5.97	%
Average Original FICO Score		710			724			—		—		720
Average Original Loan-to-Value Ratio		72.96			72.64			—		—		72.73
Average Original Combined Loan-to-Value Ratio		75.95			77.19			—		—		76.85
Underwritten with low or stated income documentation	$	44,884		$	154,981		$	—	$	—	$	199,865
Underwritten at a start rate (i.e., less than the fully indexed rate)	$	—		$	—		$	—	$	—		—
Borrowers electing to make less than full interest payments ($)	$	75,129		$	194,525		$	—	$	—	$	269,654
Borrowers electing to make less than full interest payments (percentage of total ARMS)		3	%		22	%		—		—		7	%
Amount of negative amortization (i.e., deferred interest) recorded as interest income during period	$	5,266		$	10,719		$	—	$	—	$	15,985

Set forth below as of September 30, 2009 [to be updated for December 31, 2009], are the amounts of interest income arising from the net negative amortization portion of loans and recognized in the periods noted (in thousands):

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 9

	Unpaid Principal Balance of		Amount of Net Negative
	Loans in Negative		Amortization recorded as interest
	Amortization At Year-End		income during period
2009	$	361,792	$	17,694
2008	$	314,961	$	14,787
2007	$	98,656	$	4,244

Set forth below are the frequencies at which the ARM loans, which totaled $3.8 billion at September 30, 2009 [to be updated for December 31, 2009], will reprice (in thousands):

($ in thousands)	September 30, 2009
Reset Frequency	# of loans		Balance		% of the total
Monthly		288	$	89,142		2	%
Semi-Annually		6,962		2,467,355		65	%
Annually		6,715		1,228,943		33	%
Total		13,965	$	3,785,440		100	%

Set forth below as of September 30, 2009 [to be updated for December 31, 2009], are the amounts of the ARM loans in our held-for-investment loan portfolio with interest rate reset dates in the periods noted As noted in the above table, loans may reset more than once over a three-year period. Accordingly, the table below may include the same loans in more than one period (in thousands):

	1st Quarter		2nd Quarter		3rd Quarter		4th Quarter
2010	$	821,281	$	1,040,208	$	1,084,376	$	1,133,040
2011	$	1,089,338	$	1,188,018	$	1,143,763	$	1,229,298
2012	$	1,220,741	$	1,455,642	$	1,467,275	$	1,449,314
Later years (1)	$	1,444,6441	$	1,487,400	$	1,515,521	$	1,505,686

(1)	Later years reflect one reset period per loan.
	The ARM loans were originated with interest rates that are intended to adjust, i.e., reset or reprice, within a range of an upper limit, or “caps,” and a lower limit, or “floor.” The loans in our portfolio have caps and floors that range from 10.9% to 2.8%, respectively, with the actual range dependent on the type of loan product actually held by a borrower.
	Generally, the higher the cap, the more likely a borrower’s monthly payment could undergo a sudden and significant increase due to an increase in the interest rate when a loan reprices. Such increases could result in the loan becoming delinquent if the borrower was not financially prepared at that time to meet the higher payment obligation. In the current lower interest rate environment, ARM loans have generally repriced downward, providing the borrower with a lower monthly payment rather than a higher one. As such, these loans would not have a material change in their likelihood of default due to repricing.

     Interest Only Mortgages

     Both adjustable and fixed term loans were offered with a 10-year interest only option. These loans were originated using Fannie Mae and Freddie Mac guidelines as a base framework. We generally applied the debt-to-income ratio guidelines and documentation using the AUS Approve/Accept response requirements. The LTV and CLTV maximum ratios allowable were each 100%, but subordinate financing was not allowed over a 95% LTV ratio. At a 100% LTV ratio, the FICO floor was 700, and at lower LTV levels, the FICO floor was 620. All occupancy and purpose types were allowed at lower LTVs. Lower LTV and high FICO ARMs were underwritten at the start rate, while other ARMs were either underwritten at the note rate if the initial fixed term was two years or greater, and the note rate plus two percentage points if the initial fixed rate term was six months to one year.

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 10

Set forth below is a table describing the characteristics of the interest only mortgage loans at [September 30, 2009] [to be updated at December 31, 2009] in our held-for-investment portfolio, by year of origination.

	Prior to
($ in thousands)	2007			2007			2008			2009			Total
Unpaid Principal Balance	$	1,818,334		$	925,280		$	22,072		$	1,150		$	2,766,835
Average Note Rate		5.40	%		6.22	%		6.39	%		4.95	%		5.68	%
Average Original FICO Score		723			721			749			650			722
Average Original Loan-to-Value Ratio		74.20			75.05			78.97			70.61			74.52
Average Original Combined Loan-to-Value Ratio		79.27			78.33			79.44			70.61			78.96
Underwritten with low or stated income documentation		36	%		48	%		22	%		0	%		43	%

     Second Mortgages

     The majority of second mortgages we originated were closed in conjunction with the closing of the first mortgages originated by us. We generally required the same levels of documentation and ratios as with our first mortgages. For second mortgages closed in conjunction with a first mortgage loan that was not being originated by Flagstar, our allowable debt-to-income ratios for approval of the second mortgages were capped at 40%-45%. In the case of a loan closing in which full documentation was required and the loan was being used to acquire the borrower’s primary residence, we allowed a CLTV ratio of up to 100%; for similar loans that also contained higher risk elements, we limited the maximum CLTV to 90%. FICO floors ranged from 620 to 720, and fixed and adjustable rate loans were available with terms ranging from five to 20 years.

Set forth below is a table describing the characteristics of the second mortgage loans at [September 30, 2009] [to be updated at December 31, 2009] in our held-for-investment portfolio, by year of origination.

	Prior to
($ in thousands)	2007			2007			2008			2009			Total
Unpaid Principal Balance	$	38,507		$	177,288		$	17,699		$	1,602		$	235,096
Average Note Rate		8.05	%		8.58	%		8.07	%		6.96	%		8.44	%
Average Original FICO Score		717			735			753			712			733
Average Original Loan-to-Value Ratio		21.80			19.65			18.84			17.37			19.92
Average Original Combined Loan-to-Value Ratio		89.53			90.23			78.78			93.80			89.28

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 11

     HELOCs

     The majority of home equity lines of credit (“HELOCs”) loans were closed in conjunction with the closing of related Flagstar-originated first mortgages. Documentation requirements for HELOC applications were generally the same as those required of borrowers for the Flagstar first mortgages, and debt-to-income ratios were capped at 50%. For HELOCs closed in conjunction with the closing of a first mortgage loan that was not being originated by Flagstar, our debt-to-income ratio requirements were capped at 40-45% and the LTV was capped at 80%. The qualifying payment varied over time and included terms such as either 0.75% of the line amount or the interest only payment due on the full line based on the current rate plus 0.5%. HELOCs were available in conjunction with primary residence transactions that required full documentation, and the borrower was allowed a CLTV ratio of up to 100%.; for similar loans that also contained higher risk elements, we limited the maximum CLTV to 90%. FICO floors ranged from 620 to 720. The HELOC terms called for monthly interest-only payments with a balloon principal payment due at the end of 10 years. At times, initial teaser rates were offered for the first three months.

Set forth below is a table describing the characteristics of the HELOCs at [September 30, 2009] [to be updated at December 31, 2009] in our held-for-investment portfolio, by year of origination.

	Prior to
($ in thousands)	2007			2007			2008			2009		Total
Unpaid Principal Balance	$	165,319		$	126,691		$	24,488		$	—	$	316,498
Average Note Rate		5.36	%		5.58	%		4.15	%		—		5.35	%
Average Original FICO Score		731			739			756			—		739
Average Original Loan-to-Value Ratio		25.07			24.24			23.77			—		24.88
Average Original Combined Loan-to-Value Ratio		79.74			82.50			74.45			—		80.86

     Commercial Loans

     While our primary product is single-family residential first mortgage loans originated or acquired by our home lending operation, our banking operation has in the past offered consumer and commercial financial loan products and services to individuals and to small to middle market businesses.

Our commercial loan portfolio is primarily comprised of commercial real estate (CRE) loans that are collateralized by real estate properties intended to be income-producing in the normal course of business. We mitigate our

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 12

risk on these loans by requiring collateral values that exceed the loan amount and by underwriting the loan with required cash flows in excess of the debt service requirement. These loans are made to finance properties and are repaid through cash flows related to the operation, sale, or refinancing of the property.

During 2006 and 2007, we placed an increased emphasis on commercial real estate lending and on the expansion of our commercial lending business as a diversification from our national residential mortgage lending platform. During 2008 and 2009, as a result of continued economic concerns, we funded commercial loans that had previously been underwritten and approved but otherwise halted new commercial lending activity.

The primary factors considered in past commercial credit approvals were the financial strength of the borrower, assessment of the borrower’s management capabilities, industry sector trends, type of exposure, transaction structure, and the general economic outlook. Commercial loans were made on a secured, or in limited cases, on an unsecured basis, with a vast majority also being collateralized by personal guarantees of the principals of the borrowing business. Assets used as collateral for secured commercial loans required an appraised value sufficient to satisfy our loan-to-value ratio requirements. We also generally required a minimum debt-service-coverage ratio, other than for development loans, and considered the enforceability and collectability of any relevant guarantees and the quality of the collateral.

At [September 30, 2009] [to be updated at December 31, 2009], our commercial real estate loan portfolio totaled $1.7 billion, or 22.1% of our investment loan portfolio, and our non-real estate commercial loan portfolio was $13.8 million, or 0.2% of our investment loan portfolio. At December 31, 2008, our commercial real estate loan portfolio totaled $1.8 billion, or 19.6% of our investment loan portfolio, and our non-real estate commercial loan portfolio was $24.7 million, or 0.3% of our investment loan portfolio. During 2009, we originated $29.9 million of commercial loans versus $206.0 million in 2008.

At [September 30, 2009] [to be updated at December 31, 2009], our commercial real estate loans were geographically concentrated in a few states with approximately $0.9 billion or 53.9% of all commercial loans located in Michigan, $238.1 million (13.9% ) located in Georgia and $166.9 million (9.7%) located in California.

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 13

In commercial lending, ongoing credit management is dependent upon the type and nature of the loan. We monitor all significant exposures on a regular basis. Internal risk ratings are assigned at the time of each loan approval and are assessed and updated with each monitoring event. The frequency of the monitoring event is dependent upon the size and complexity of the individual credit, but in no case less frequently than every 12 months. Current commercial collateral values are updated more frequently if deemed necessary as a result of impairments of specific loan or other credit or borrower specific issues. We continually review and adjust our risk rating criteria and rating determination process based on actual experience. This review and analysis process also contributes to the determination of an appropriate ALLL amount for our commercial loan portfolio.

The following table identifies our commercial loan portfolio by major category and selected criteria at [September 30, 2009] [to be updated at December 31, 2009]:

As of September 30, 2009 [update at December 31, 2009]

	Unpaid
	Principal		Average			Commercial Loans on Non-
($ in thousands)	Balance		Note Rate			accrual Status
Commercial Real Estate loans:
Fixed rate	$	841,608		7.2	%	$	159,262
Adjustable rate		844,084		8.9	%		254,654
Total commercial real estate	$	1,685,692		8.1	%	$	413,916
Commercial Non-Real Estate loans:
Fixed rate	$	6,774		6.4	%	$	805
Adjustable rate		7,032		6.7	%		5,081
Total commercial non-real estate	$	13,806		6.6	%	$	5,886

We also continue to offer warehouse lines of credit to other mortgage lenders. These commercial lines allow the lender to fund the closing of a residential mortgage loan. Each extension or drawdown on the line is collateralized by the residential mortgage loan being funded, and in many cases, we subsequently acquire that loan. Underlying mortgage loans must be originated based on our underwriting standards. These lines of credit are, in most cases, personally guaranteed by one or more qualified principal officers of the borrower. The aggregate amount of warehouse lines of credit granted to other mortgage lenders at [September 30, 2009] [to be updated at December 31, 2009], was $1.1 billion, of which $425.9 million was outstanding, as compared to, $1.1 billion granted at December 31, 2008, of which $434.1 million was outstanding.

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 14

Comment No. 1(c):	We note that you include the amount of negative amortization recorded as interest income in a table by year of origination. It is not clear from this presentation when the interest income was recognized. Please revise to specifically disclose this amount for each period reported in your financial statements. Consider the need to specifically discuss these amounts in MD&A.
Response No. 1(c):	In response to the Staff’s comment, set forth below as of September 30, 2009 is a tabular presentation of the interest income recognized by us arising from negative amortization, by period recognized. We will include this presentation in our future filings.
	“Set forth below as of September 30, 2009 [to be updated for December 31, 2009], are the amounts of interest income arising from the negative amortization portion of loans and recognized in the periods noted (in thousands):

	Unpaid Principal Balance of		Amount of Net Negative
	Loans in Negative		Amortization recorded as interest
	Amortization At Year-End		income during period
2009	$	361,792	$	17,694
2008	$	314,961	$	14,787
2007	$	98,656	$	4,244

Further, in response to the Staff’s comment, we will include disclosure in our MD&A, under Results of Operations — Net Interest Income,” as follows:

“Our interest income also includes the amount of negative amortization (i.e., deferred interest) arising from our option ARM loans. See “Item 1 — Business — Operating Segments — Home Lending Operation — Underwriting.” The amount of net negative amortization included in our interest income during [nine months ended September 30, 2009] [ to be updated for full year 2009], 2008 and 2007 were $17.7 million, $14.8 million and $4.2 million, respectively.”

Note 8. Private-label Securitization Activity, page 97

Comment No. 2: Please refer your response to comment 2 of our letter dated September 30, 2009 and address the following:

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 15

Comment No. 2(a):	Please revise future filings to incorporate your response into your future filings. For example, discuss the value of the transferor’s interest of each trust separately, and explain your views as to why you believe the amounts recorded will be collected.
Response No. 2(a):	In future filings, we will incorporate our response to number 2 of your letter dated September 23, 2009, as appropriate. In addition, we will include the following,
	FSTAR 2005-1 With respect to this securitization, Flagstar carried a residual interest of $2.1 million as of November 30, 2009. At the same time, Flagstar had a carrying value of $19.1 million in transferor’s interests. However, we did not view these transferor’s interests as being at risk of loss because (i) there were no outstanding claims owed to the note insurer at that time, (ii) we continued to receive cash flows on the interest payments associated with the transferor’s interest, (iii) increases in transferor’s interests gave rise to increases in the cash inflow into the securitization trust that thereby improved the relative credit positions of all parties to the securitization, and (iv) the structure of the securitization provided for losses in the transaction to be shared equally, i.e.,pari passu, among the parties rather than being borne solely or primarily by Flagstar . As such, we determined that a SFAS 5 liability (now codified within ASC Topic 450, “Contingencies,”) did not arise and therefore a recording of a liability was not warranted.
	FSTAR 2006-2 With respect to this securitization, as of November 30, 2009, the residual interests had a fair value of $0. The fair value of the residual interest had been written down to $0 since the third quarter of 2008. At November 30, 2009, outstanding claims due to the note insurer were $39.9 million and based on our internal model, we believed that because of the claims due to the note insurer and continuing credit losses on the loans underlying the securitization, the carrying amount of the transferor’s interest was $0. Also, during the fourth quarter 2009, Flagstar determined that the transferor’s interests had deteriorated to the extent that a SFAS 5 (now codified within ASC Topic 450, “Contingencies,”) liability was required to be recorded. During the period, Flagstar recorded a liability of $7.6 million to reflect the expected liability arising from losses on future draws associated with this securitization. In determining this liability, Flagstar (i) assumed no further draws would be made with respect to those HELOCs as to which further draws were currently prohibited, (ii) the remaining HELOCs would continue to operate in the same manner as their historical draw behavior indicated, as measured on an individual loan basis and on a pool drawdown basis, and (iii) that any draws actually made and therefore recognized as transferor’s interests by

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 16

	Flagstar would have a loss rate of 100%. Flagstar expects to review the amount of this liability at least quarterly and adjust it accordingly.
Comment No. 2(b):	Please tell us and revise future filings to separately disclose the notional value of the total unfunded commitment under each trust and the amount of the unfunded commitments that have been frozen or suspended under each trust. As part of your disclosure, discuss how you developed the SFAS 5 (ASC 450) liability recorded for the trust(s) and disclose the key assumptions, such as the fact that you are assuming a 100% loss on all of the subsequent draw downs on the HELOCs and how you developed the estimate based on historical draw behaviors. Additionally, supplementally tell us in greater detail how you estimated the draw downs based on historical behavior, and consider providing us an example to illustrate how the amount was determined.
Response No. 2(b):	As part of our future filings, we will disclose the following information for each of our HELOC trusts as shown below:
	The table below identifies separately for each HELOC trust: (i) the notional amount of the unfunded commitment under the Company’s contractual arrangements, (ii) unfunded commitments that have been frozen or suspended because the borrowers do not currently meet the contractual requirements under their home equity line of credit with the Company, and (iii) the amount currently fundable because the underlying borrowers’ lines of credit are still active:

As of November 30, 2009

($ in thousands)	FSTAR 2005-1		FSTAR 2006-2		Total
Notional Amount of Unfunded Commitments (1)	$	45,290	$	42,597	$	87,887
Frozen or Suspended Unfunded Commitments	$	20,571	$	24,750	$	45,341
Unfunded Commitments Still Active	$	24,719	$	17,846	$	42,565

(1) The Company’s total potential funding obligation is dependent on both (a) borrower behavior (e.g., the amount of additional draws requested) and (b) the contractual draw period (remaining term) available to the borrowers. Because borrowers can make principal payments and restore the amounts available for draws and then borrow additional amounts as long as their lines of credit remain active, the funding obligation has no specific limitation and it is not possible to define the maximum funding obligation. However, we expect that the call provision of this securitization pool will be reached in 2015 and our exposure will be substantially mitigated at that time, based on prepayment speeds and losses in our cash flow forecast.

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 17

As previously discussed, we do not believe that a liability has been incurred for expected losses on future draws under FSTAR 2005-1. The securitization continues to have significant value in its residual and transferor’s interest and the note insurer has not been required to perform under its contract.

During the fourth quarter 2009, we determined that a liability for expected losses on future draws was required to be recorded for FSTAR 2006-2 as the value of our transferor interest approached $0. In order to estimate losses on future draws and the timing of such losses, a forecast for the draw reserve was established. The forecast was used as the basis for recording the liability. Historical observations and draw behavior formed the basis for establishing the key assumptions and forecasted draw reserve. First, the forecast assumed a 100% loss on all future draws. Second, the forecast projected future obligations on a monthly basis using a three-month rolling average of the actual draws as a percentage of the unfunded balance.

For example, for the period ended November 30, 2009, the three- month rolling average draw rate was 2.98% of the unfunded commitments(still active). This percentage was computed by dividing (i) the actual draw rate over the three month period ending on that date, by (ii) the balance of the unfunded commitments still active on that date. With respect to November 2009, such three month period ended on November 30, 2009.

The draw rate was then used to project monthly draws through the remaining expected life of the securitization. In doing so, the 2.98% draw rate (as noted above) was applied against the expected declining level of unfunded commitments in future months caused by payoffs, credit terminations and line cancellations, This rate of decline was based on historical experience within the securitization pool of loans.

These calculations of future monthly draws comprise, in the aggregate, the total dollar amount of expected future draws from the securitization pool. Despite a significant reduction in the unfunded commitments, we have not observed a similar reduction in the actual draw rate. Even with a constant draw rate, such total dollar amount declines to the extent the level of unfunded commitments that are still active declines, as is the case in our forecast. Nonetheless, the rate of draw remains constant, which is reflected in the constant rate of expected total future draws of 42.45% (computed as dollar amount of expected future draws divided by dollar amount of now-current dollar amount of unfunded commitments still active).

Because the expected loss on future draws in November 2009 was 100%, the expected future draws equaled the potential future draw liability at that date. We used the same analysis as of June 30, 2009 and September 30, 2009; however, because the expected loss on future draws was less than 100% on each of those dates, the expected future draws determined as of such dates were each multiplied by their respective expected loss rate to determine the potential future liability on such date.

As indicated above, the forecast uses a constant draw rate as a percentage of the current unfunded balance that is based on historical observations and draw behavior. The forecast does not contemplate currently inactive accounts becoming active and thereby becoming eligible for draw. This is because the nature of the loans that do not currently generate transferor’s interests have characteristics that suggest an extremely low likelihood of doing so in the future. Such loans are those in which the draw feature has been discontinued pursuant to the terms of the underlying loan agreement due to a credit-related deficiency of the borrower or due to a decline in the value of the related residential property serving as collateral.

The forecast also reflects the low or zero draw rates of certain of the unfunded commitments that are still active (i.e., $24.7 million for FSTAR 2005-1 and $17.8 million for FSTAR 2006-2 at November 30, 2009). For instance, some loans are still active but have never been drawn upon, suggesting that the loan may have been acquired at the time of a related first mortgage origination solely for contingency purposes but without any actual intent to draw. Similarly, another group of active loans were fully drawn upon at the time of the related first mortgage origination and have been paid down over time, suggesting that the borrower intended the HELOC to serve more as a second mortgage rather than as a revolving line of credit.

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 18

Comment No. 2(c):

It is unclear to us how you concluded that a SFAS 5 (ASC 450) liability was not required for the FSTAR 2006-2 trust prior to November 30, 2009 given the losses recognized on the transferor’s interest as of June 30, 2009 and September 30, 2009. Please either provide further information supporting your conclusion that a SFAS 5 (ASC 450) liability was not required in earlier periods or provide us with a materiality analysis illustrating how the effect would not have been material.

Response No. 2(c):

Our principal reason for not recording a SFAS 5(ASC 450) liability related to losses expected on future draws with respect to FSTAR 2006-2 is because we did not believe that the expected losses were material to our financial statements. The following table outlines our expected losses on future draws at each period indicated.

FSTAR 2006-2

			Expected Future
			Draws as % of			Expected					Potential
	Unfunded		Unfunded			Future		Expected			Future
	Commitments(1)		Commitments(2)			Draws(3)		Loss(4)			Liability(5)
June 30, 2009	$	35,499,131		42.45	%	$	15,068,159		22.75	%	$	3,428,006
September 30, 2009	$	21,773,682		42.45	%	$	9,242,178		89.18	%	$	8,242,175
November 30, 2009	$	17,846,201		42.45	%	$	7,575,098		100.00	%	$	7,575,098

(1)	Unfunded commitments represent the amounts currently fundable at the dates indicated because the underlying borrowers’ lines of credit are still active..
(2)	Expected future draws on unfunded commitments represents the historical draw rate within the securitization . The historical draw rate has been consistent during 2009
(3)	Expected future draws reflects unfunded commitments multiplied by expected future draws percentage.
(4)	Expected losses represents an estimated reduction in carrying value of future draws.
(5)	Potential future liability reflects expected future draws multiplied by expected losses.

There are two distinct components to the assumptions underlying the loss rate on the transferor’s interests. First, the structure of the securitization provided for losses in the transaction to be sharedpari passu, i.e., equally, among the parties rather than being borne solely or primarily by Flagstar. Second, to the extent that underlying claims to the insurer increased concurrently with credit losses, the reimbursement owed to the insurer from the waterfall also increased. In June, we projected that the excess spread was sufficient to cover all expected claims due to the insurer. At that time, the loss rate for the transferor’s interests represented the present value of the insurer’s share in the collateral pool’s forecasted losses. During the quarter ended September 30, 2009, our projected cumulative credit losses for loans in the securitization increased significantly as previously disclosed in the Private-Label Securitization Activity footnote to our previously filed Form 10-Q’s for 2009. Based on this significant increase , we projected in September that the insurer’s claims were no longer supported solely by the excess spread, but instead a portion of the transferor’s interest was also required in order to reimburse the insurer for its claims. Consequently, the loss rate for September included both the share of the

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 19

collateral pool’s forecasted losses and the amount of the insurer’s claims not reimbursed through excess spread. As such, we determined that there was a significant increase to the expected loss rate for future draws at September 30, 2009. However, after reviewing the effect of the revised loss rate and the resulting potential future liability, we determined, and our auditors concurred, that our recording of a liability regarding transferors’ interests at that time would not have a material effect on our financials statements. However, in November, the excess spread and the transferor’s interests were insufficient to support the repayment of the insurer’s claims, and the assumed loss rate increased to 100%, giving rise to our recording of the related liability at that time.

Based on pre-tax losses of $102.9 million and $178.6 million for the three month periods ended June 30, 2009 and September 30, 2009, respectively, we do not believe that the fact that the estimated liability was not recorded earlier would have had a material impact on our financial statements for either period on a quantitative or qualitative basis. In each case, the monetary difference was less than 5% of pre-tax loss or stockholder’s equity. Further, on a qualitative basis, we do not believe that either the absence of such a liability or recording such liability in the three-month periods ended June 30, 2009 or September 30, 2009 would have been significant to an investor. Among other things, we do not believe that the nature and amount of the change, (1) would have masked a change in earnings or other trends, (2) would have affected our ability or inability to meet analysts’ expectations, (3) would have changed a loss to earnings or vice versa, (4) would have a significant effect on segment information, (5) would have affected our compliance with regulatory requirements or significantly affected our capital levels, (6) would have affected management bonuses in any significant way, or (7) would have involved the concealment of an unlawful transaction.

Form 8-K filed December 14, 2009

Comment No. 3 Through Comment No. 6: As discussed with the Staff, we have reviewed Comments 3 through 6 and agree with them. As disclosed in the draft Pre-Effective Amendment No. 1 to Form S-3 (the “Form S-3”), we continue to have discussions with our regulators and we have disclosed what we believe are the current material regulatory restrictions applicable to us that are now disclosable. When these matters are definitively resolved and disclosable, we

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 20

undertake to update the market and the Form S-3, once filed, through the filing of a new Form 8-K.

Form S-3 filed November 3, 2009

Comment No. 7:	Please revise the Form S-3 that you filed with Commission on November 3, 2009 to include the information provided in your responses to our comments. Alternatively, amend the affected documents or provide the information in a Form 8-K to be specifically incorporated into the registration statement.
Response No. 7:	As discussed with the Staff, we will include the information provided in our responses to the Staff’s comment in a Form 8-K to be incorporated by reference into the Form S-3.
Comment No. 8:	File a pre-effective amendment to the Form S-3 filed on November 3, 2009 that includes the prospectus supplement for your proposed rights offering. Include a complete and prominent description of the additional regulatory restrictions recently imposed on the company and include related risk factors, as appropriate.
Response No. 8:	As discussed with the Staff, we will file a pre-effective amendment to the Form S-3 that includes the prospectus supplement for the proposed rights offering. The prospectus supplement will include a discussion of the regulatory restrictions in the prospectus summary and any relevant risk factors.

*     *     *     *     *

Mr. Christian Windsor
Securities and Exchange Commission
December 30, 2009
Page 21

     Flagstar Bancorp, Inc. hereby acknowledges that:

• it is responsible for the adequacy and accuracy of the disclosure in the filings;

• staff comments or changes to disclosure in response to staff comments do not foreclose the Commission from taking any action with respect to the filing; and

• it may not assert staff comments as a defense in any proceeding initiated by the Commission or any person under the federal securities laws of the United States.

     We trust that the forgoing is responsive to your comments. If you have any questions regarding the foregoing or require further information, please contact the undersigned at (248) 312-5580.

Sincerely, FLAGSTAR BANCORP, INC.
	/s/ Paul. D. Borja
	By:	Paul D. Borja
	Its: Executive Vice President and Chief Financial Officer