September 16, 2008


VIA FACSIMILE TRANSMISSION - 202-772-9209
Securities and Exchange Commission
Division of Corporation Finance
450 Fifth Street, N.W.
Washington, D.C. 20549

Attn:
William H. Demarest IV
Staff Accountant
Division of Corporate Finance
 
Re:
Redwood Trust, Inc.
Responses to Comments on Form 10-K and Schedule 14A
Filed on March 5, 2008 and April 22, 2008, respectively
File No. 1-13759

 
Dear Mr. Demarest,

On behalf of Redwood Trust, Inc. (“Redwood”), I hereby provide the following responses in reply to the Staff’s comment letter dated September, 3, 2008 (the “Comment Letter”) in connection with the above-referenced Annual Report on Form 10-K and Definitive Proxy Statement on Schedule 14A.

The responses to the Staff’s comments are numbered to relate to the corresponding comments in the Comment Letter. For your convenience, each of our responses is preceded with an italicized recitation of the corresponding comment set forth in the Comment Letter.

Form 10-K for the year ended December 31, 2007

Item 9A. Controls and Procedures

Management’s Report on Internal Control over Financial Reporting, page 90

 
1.
We note that you have concluded that your internal controls over financial reporting were ineffective as of December 31, 2007 because you were unable to obtain the necessary evidence to support your initial conclusion that unrealized losses are recoverable. Please clarify to us how all such unrealized losses have been accounted for in your audited financial statements.

As part of our financial reporting process, we are required to assess whether declines in the fair value of available-for-sale (AFS) securities below their amortized cost should be reflected as other-than-temporary (OTT) impairments. The diagram below highlights the three step process we follow.


  

Our internal controls, processes, and assessments for OTT impairments were appropriate under steps one and two shown above. It was during our evaluation under step three, which is highly subjective, that in certain circumstances we were unable to obtain the necessary evidence as prescribed under SAB 59 to support our initial conclusions that these securities were not other-than-temporarily impaired. As a consequence, we determined that our internal controls over financial reporting as applied to this specific process were ineffective. We have since made enhancements to our internal controls relating to this process to improve their effectiveness going forward.

In all circumstances where we were unable to obtain sufficient evidence under SAB 59, we treated all of those AFS securities as other-than-temporarily impaired. We recorded these impairments as negative market valuation adjustments, which were recognized in our income statement for the quarter and year ended December 31, 2007.
 
Note 6. Real Estate Securities, page F-26

 
2.
We note that you have recorded losses in the amount of $132 million for real estate securities that have been in a loss position for greater than twelve months. Please tell us what evidence exists to support the recoverability of the fair value of these securities and why the losses have not been recognized into earnings. Refer to SAB Topic 5:M.
 
The table below summarizes the results of our impairment analysis on AFS securities as of December 31, 2007. This summary includes the $132 million of unrealized losses for securities that have been in a loss position for greater than twelve months, and also illustrates the significant amounts of OTT impairment recognized in our income statement for the quarter ended December 31, 2007.
 

Impairment Summary - Q4 2007

   
Amount of Impairment
in millions
  
< 12 months
  
> 12 months
  
Total
  
Impairments on AFS Securities at 12/31/2007
  
$
(1,357
)
$
(273
)
$
(1,630
)
Other-Than-Temporary (OTT) Impairment Test:
                    
Step 1: Adverse Cash Flows
    
(268
)
  
(28
)
  
(296
)
Step 2: Intent and Ability to Hold
    
-
    
-
    
-
  
Step 3: Recoverability
    
(660
)
  
(113
)
  
(773
)
Impairments Deemed OTT
  
$
(928
)
$
(141
)
$
(1,069
)
                           
Remaining Unrealized Losses
  
$
(429
)
$
(132
)
$
(561
)
 
As noted above, we employ a systematic three step testing process to determine whether an impairment is other-than-temporary and we consider all available information, positive and negative. The third and final step in this process is to consider the recoverability of the realizable value of the security in a reasonable amount of time in light of market conditions.

For impaired securities (especially those that have been in a loss position for greater than twelve months), we considered the following evidence to support our conclusion that an OTT impairment had not occurred.

·
The length of time that fair value has been less than cost:

The AFS securities we own are real estate securities, most of which have below investment grade (rated BB, B, or nonrated) credit ratings. These securities are illiquid in nature with a limited secondary market. Although GAAP disclosures distinguish between securities that have had unrealized losses for periods of more than and less than twelve months, we share the SEC staff’s view that this one-year time period is not an automatic line of demarcation for inferring when unrealized losses become other-than-temporary impairments. As evidenced historically, we believe a reasonable period of time to observe a recovery in value for illiquid real estate securities is often a longer period than one year. We intend to hold these securities through their longer, yet reasonable, forecasted recovery periods.

As a REIT established in 1994, we act as a long term investor with the intent and ability to hold investment securities through their recovery periods. We fund AFS securities with equity or through bankruptcy remote securitization entities and we have very little recourse debt. At December 31, 2007, we had $8 million of short-term recourse debt and $290 million in unrestricted cash.
 
·
The extent to which fair value has been less than cost:

We believe each security’s fair value reflected the market's most recent evaluation of the total mix of available information, despite historically low liquidity and pricing transparency for real estate securities. As part of our testing process, we considered the severity of the decline in fair value relative to the amortized cost basis of each security. As this severity increased, we weighted it higher in our overall assessment of recoverability.

In reviewing the severities of declines in fair values, we also considered the trending of bid/ask spreads required by market makers for these illiquid real estate securities. These spreads were at historical highs at the time we filed our Form 10-K, contributing to a significant portion of the decline in fair values we observed during the second half of 2007. A narrowing of this bid/ask spread due to an increase in the number of active market participants should help to recover much of these unrealized losses for temporarily impaired securities.
 
 
Our review generally resulted in OTT impairments on securities with significant declines in fair value, with final determination based upon facts and circumstances specific to each security. This included $113 million of OTT impairments on securities with unrealized losses for periods of more than twelve months. The remaining securities generally had less severe declines in fair value and strong cash flows, which allowed us to conclude that they should recover in value.

·
Characteristics specific to each security:
 
We considered the characteristics specific to each impaired security and the general health of the market for that security as part of our assessment of recoverability. We observed a general deterioration in underwriting quality for residential loans originated during 2006 and 2007, which collateralize some of our real estate securities. Specifically, loan-to-value ratios, debt-service ratios, and borrower representations had become very aggressive relative to prior vintages. Loans originated prior to 2006 generally exhibited less aggressive underwriting standards and benefited from lower leverage due in part to significant home price appreciation. The CDO securities we own are complex structured products that are primarily backed by residential real estate loans and securities.

As the following table shows, our assessment process resulted in OTT impairment on most residential securities originated during 2006 and 2007 and all CDO securities. We also recognized OTT impairment on a significant portion of non-prime residential securities originated in 2005 and prior years.
 
Summary of Impairments> 12 months at 12/31/2007
(in millions)
               
Type of Impairment
  
 Other-Than-Temporary
  
Temporary
  
Total Impairments > 12 months
  
By Vintage
  
2000-2005
 
2006-2007
 
Total
 
2000-2005
 
2006-2007
  
Total
  
2000-2005
  
2006-2007
  
Total
  
Real Estate Securities:
                                                        
Residential Prime
  
$
-
  
$
(10
)
$
(10
)
$
(54
)
$
(12
)
$
(66
)
$
(54
)
$
(22
)
$
(76
)
Residential Non-Prime
    
(28
)
  
(83
)
  
(111
)
  
(27
)
  
(2
)
  
(29
)
$
(55
)
$
(85
)
$
(140
)
CDO
    
(15
)
  
(4
)
  
(19
)
  
-
    
-
    
-
  
$
(15
)
$
(4
)
$
(19
)
Commercial
    
-
    
(1
)
  
(1
)
  
(27
)
  
(10
)
  
(37
)
$
(27
)
$
(11
)
$
(38
)
Total
  
$
(43
)
$
(98
)
$
(141
)
$
(108
)
$
(24
)
$
(132
)
$
(151
)
$
(122
)
$
(273
)

Of our securities with unrealized losses greater than 12 months at December 31, 2007, a significant portion consisted of residential mortgage-backed securities (RMBS) collateralized with prime and non-prime residential loans originated in 2005 and prior years. They were generally purchased at significant discounts and are sensitive to small changes in valuation assumptions (e.g., prepayment speeds) in the near-term. We determined that the performance of the underlying loans to date, the market’s outlook for future performance, and the projected receipt of principal and interest provided us with sufficient support to anticipate that fair values on these specific securities would recover within a reasonable period of time.

The other significant portion of our securities with unrealized losses greater than 12 months at December 31, 2007, consisted of commercial mortgage-backed securities (CMBS). These securities are the lower rated securities within a commercial mortgage securitization and most are backed by loans originated in 2005 and prior years. Over the prior six months, we saw an increase in yields required by the market for CMBS in line with other structured real estate securities. However, the CMBS market seemed to be healthy with new securitizations planned at that time and little evidence that the cash flows on underlying commercial loans were deteriorating. Given the general health of the commercial market, the relatively short period of price declines, and the strong cash flows on the underlying mortgages, we determined that there was sufficient evidence to support our assessment that fair values would recover within a reasonable period of time.
 
 
·
The financial condition and near-term prospects of the issuer:

Since these securities are sold by bankruptcy remote securitization trusts, the financial concerns are limited to the collateral credit performance and not the financial condition and near term prospects of the issuer. As noted above, there are several steps involved in our assessment of impairments, including the determination of whether there has been an adverse change in cash flows. The securities we deemed not OTT impaired have demonstrated strong credit performance, which supports our assessment that these securities will recover within a reasonable period of time.

·
The anticipated recovery period for securities not OTT impaired:
 
During the second half of 2007, prices of certain mortgage-backed securities declined precipitously in a relatively short period of time. The following factors were considered prior to our filing of Form 10-K (using professional judgments based upon information available at the time) in our determination that the value of certain securities was recoverable in a reasonable period of time:

 
Ø
There were two times in the past ten years (1998 and 2002-2004) when interest spreads significantly widened for real estate securities (prices declined) on a systematic basis. Following both of these periods, spreads tightened back to historical levels. Available information led us to conclude this cyclical process would continue and promote future recoveries in fair value.

 
Ø
There had been significant contraction in available liquidity and the extension of credit. This led to distressed selling by many hedge funds and other market participants who used financial leverage to acquire securities. Available information led us to conclude that the capital markets would calm and more orderly trading would resume over time, allowing fair values to recover.

 
Ø
The contraction in the extension of credit by banks kept many potential buyers out of the market as they could not obtain financing for new investments. Available information led us to conclude that the efforts of the Federal Reserve in reducing the discount rate and accepting more forms of collateral at the Fed window would make credit more available to borrowers, which would benefit fair values by promoting market liquidity.

 
Ø
Historically low prepayments had significantly affected the fair values of many real estate securities. Available information led us to conclude that Fed interest rate cuts, economic stimulus plans, and other initiatives undertaken by the United States government would result in higher mortgage prepayments going forward and promote a recovery in fair values of real estate securities purchased at discounts.

After following the three step methodology and assessing the results in accordance with SAB Topic 5: M, we determined that sufficient evidence existed for us to conclude that securities that had been temporarily impaired for more than twelve months would recover, in a reasonable amount of time, the aggregate unrealized losses in the amount of $132 million. We therefore concluded that these unrealized losses were not OTT impairments at December 31, 2007.

 
Subsequent market developments:

As of December 31, 2007 (and through our filing date of March 5, 2008), the extent and severity of the housing correction was less clear than it is today. Since we filed our Annual Report on Form 10K, there have been subsequent events affecting both the RMBS and CMBS markets. These include events such as the acquisition of Bear Stearns by JP Morgan Chase, increased margin calls on highly leveraged firms (including REITs like Thornburg Mortgage), increased and continuing writedowns on the values of securities and loans for many financial institutions, the FDIC takeover of IndyMac, the federal government placing Fannie Mae and Freddie Mac into conservatorship, and Lehman Brothers filing for bankruptcy. In addition, housing prices have continued to fall in most markets and the trend has been generally worse than previously expected. Increases in unemployment and continued slow economic growth have also contributed to increases in residential mortgage delinquencies and deteriorating performance on commercial properties.

As these events unfold, we continue to apply our assessment process at every reporting period and reassess whether securities with unrealized losses should deemed other-than-temporarily impaired.
 
Exhibit 31

 
3.
We note that the certifications are not in the proper form. Specifically, you have omitted the parenthetical disclosure from paragraph 4(d). The required certifications must be in the exact form prescribed; the wording of the required certifications may not be changed in any respect. In future filings please ensure that the certifications of your current Principal Executive Officer and Principal Financial Officer are in the exact form currently set forth in Item 601(b) (31) of Regulation S-K.

In future filings, we will ensure that the certifications of our current Principal Executive Officer and Principal Financial Officer are in the exact form set forth in Item 601(b)(31) of Regulation S-K.

Schedule 14A

Base Salary, page 15

 
4.
In future filings please provide a more detailed analysis of the reasons for any material changes in named executive officer base salaries. For example, we note that in 2007 almost all named executive officers received material increases in base salary.

In future filings, we will provide a more detailed analysis of the reasons for any material changes in named executive officer base salaries.

Performance-Based Compensation, page 15

 
5.
In future filings, with respect to individual performance awards, please provide a more detailed analysis of how the company determined the awards for each named executive officer. Please discuss how the factors, which you list on the bottom of page 15, were analyzed and how such analysis resulted in the actual payment amounts.

In future filings, we will provide more detailed analysis of how the company determined the awards for each named executive officer. The discussion will include how the factors we list were analyzed and how such analysis resulted in the actual payout amounts.
 
 
As you have requested, this will confirm that:

 
·
Redwood Trust is responsible for the adequacy and accuracy of the disclosure in the above-referenced 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 filings; and

 
·
Redwood Trust 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.

Should you have any further comments or questions about this letter, please contact me by telephone at 415-380-3455, by fax at 415-381-1773, or by email at [email protected].
 
 
Very truly yours,
     
 
Redwood Trust, Inc.
     
     
 
By:
/s/  MARTIN S. HUGHES
   
Martin S. Hughes
   
Chief Financial Officer