9.	Loss Reserves

As described in Note 3 – "Summary of Significant Accounting Policies," we establish reserves to recognize the estimated liability for losses and loss adjustment expenses related to defaults on insured mortgage loans. Loss reserves are established by estimating the number of loans in our inventory of delinquent loans that will result in a claim payment, which is referred to as the claim rate, and further estimating the amount of the claim payment, which is referred to as claim severity.

Estimation of losses is inherently judgmental. The conditions that affect the claim rate and claim severity include the current and future state of the domestic economy, including unemployment, and the current and future strength of local housing markets. Current conditions in the housing and mortgage industries make these assumptions more volatile than they would otherwise be. The actual amount of the claim payments may be substantially different than our loss reserve estimates. Our estimates could be adversely affected by several factors, including a deterioration of regional or national economic conditions, including unemployment, leading to a reduction in borrowers' income and thus their ability to make mortgage payments, and a drop in housing values that could result in, among other things, greater losses on loans that have pool insurance, and may affect borrower willingness to continue to make mortgage payments when the value of the home is below the mortgage balance, and mitigation from rescissions being materially less than assumed. Changes to our estimates could result in a material impact to our results of operations and capital position, even in a stable economic environment.

The following table provides a reconciliation of beginning and ending loss reserves for each of the past three years:

	2012			2011			2010
	(In thousands)

Reserve at beginning of year	$	4,557,512		$	5,884,171		$	6,704,990
Less reinsurance recoverable		154,607			275,290			332,227
Net reserve at beginning of year (1)		4,402,905			5,608,881			6,372,763
Adjustment to reserves (2)		-			-			(92,000	)
Adjusted beginning reserves		4,402,905			5,608,881			6,280,763

Losses incurred:
Losses and LAE incurred in respect of default notices received in:
Current year		1,494,133			1,814,035			1,874,449
Prior years (3)		573,120			(99,328	)		(266,908	)
Subtotal (4)		2,067,253			1,714,707			1,607,541

Losses paid:
Losses and LAE paid in respect of default notices received in:
Current year		134,509			121,383			60,897
Prior years (5)		2,389,985			2,838,069			2,256,206
Reinsurance terminations (6)		(6,331	)		(38,769	)		(37,680	)
Subtotal (7)		2,518,163			2,920,683			2,279,423
Net reserve at end of year (8)		3,951,995			4,402,905			5,608,881
Plus reinsurance recoverables		104,848			154,607			275,290

Reserve at end of year	$	4,056,843		$	4,557,512		$	5,884,171

(1)

At December 31, 2011, 2010 and 2009 the estimated reduction in loss reserves related to rescissions approximated $0.7 billion, $1.3 billion and $2.1 billion, respectively.

(2)

In periods prior to 2010 an estimate of premium to be refunded in conjunction with claim payments was included in Loss Reserves. In 2010, we separately stated portions of this liability in Other liabilities and Premium deficiency reserve on the consolidated balance sheet.

(3)

A negative number for prior year losses incurred indicates a redundancy of prior year loss reserves, and a positive number for prior year losses incurred indicates a deficiency of prior year loss reserves. See table below regarding prior year loss development.

(4)

Our estimated rescissions were reduced by approximately $0.2 billion in 2012 due to our probable settlement agreements (See Note 20 – "Litigation and Contingencies"), other rescissions had no significant impact on our losses incurred in 2012. Rescissions mitigated our incurred losses by an estimated $0.2 billion in 2010. Rescissions did not have a significant impact on incurred losses in 2011.

(5)

In 2012, includes $0.1 billion payment under the Freddie Mac settlement agreement (See Note 20 – "Litigation and Contingencies").

(6)

In a termination, the reinsurance agreement is cancelled, with no future premium ceded and funds for any incurred but unpaid losses transferred to us. The transferred funds result in an increase in our investment portfolio (including cash and cash equivalents) and a decrease in net losses paid (reduction to losses incurred). In addition, there is an offsetting decrease in the reinsurance recoverable (increase in losses incurred), and thus there is no net impact to losses incurred. (See Note 11 – "Reinsurance")

(7)

Rescissions mitigated our paid losses by an estimated $0.3 billion, $0.6 billion and $1.0 billion in 2012, 2011 and 2010, respectively, which excludes amounts that may have been applied to a deductible.

(8)

At December 31, 2012, 2011 and 2010 the estimated reduction in loss reserves related to rescissions approximated $0.2 billion, $0.7 billion and $1.3 billion, respectively.

The "Losses incurred" section of the table above shows losses incurred on default notices received in the current year and in prior years. The amount of losses incurred relating to default notices received in the current year represents the estimated amount to be ultimately paid on such default notices. The amount of losses incurred relating to default notices received in prior years represents the actual claim rate and severity associated with those defaults notices resolved in the current year differing from the estimated liability at the prior year-end, as well as a re-estimation of amounts to be ultimately paid on defaults remaining in inventory from the end of the prior year. This re-estimation of the estimated claim rate and estimated severity is the result of our review of current trends in the default inventory, such as percentages of defaults that have resulted in a claim, the amount of the claims, changes in the relative level of defaults by geography and changes in average loan exposure.

Losses incurred on default notices received in the current year decreased in 2012 compared to 2011, and in 2011 compared to 2010, primarily due to a decrease in the number of new default notices received, net of cures. These factors were somewhat offset by a smaller benefit from captive arrangements.

The prior year development of the reserves in 2012, 2011 and 2010 is reflected in the table below.

	2012			2011			2010
	(In millions)
Prior year loss development:

Pool policy settlement (1)	$	267		$	-		$	-

Estimated probable rescission settlements (1)		100			-			-

Increase (decrease) in estimated claim rate on primary defaults		260			200			(432	)
Decrease in estimated severity on primary defaults		(70	)		(165	)		-
Change in estimates related to pool reserves, LAE reserves and reinsurance (2)		16			(134	)		165
Total prior year loss development	$	573		$	(99	)	$	(267	)

(1)

See Note 20 - "Litigation and Contingencies" for a discussion of our settlement with Freddie Mac and our probable settlements regarding rescissions.

(2)

Includes approximately ($114) million related to LAE reserves in 2011 and approximately $185 million related to pool reserves in 2010.

The increase (decrease) in the claim rate was based on the resolution of approximately 55%, 57% and 55% in 2012, 2011 and 2010, respectively of the prior year default inventory, as well as a re-estimation of amounts to be ultimately paid on defaults remaining in inventory from the end of the prior year and estimated incurred but not reported items from the end of the prior year. In addition, during 2012, lower estimated rescission rates, as well as our experience on defaults that are 12 months or more delinquent have increased our estimate of the claim rate. In 2010, our estimated claim rates decreased due to greater cures experienced, a portion of which resulted from loan modifications. The decrease in the estimated severity in 2012 and 2011 was based on the resolution of the prior year default inventory. The decrease in estimated loss adjustment expense in 2011was based on recent historical trends in the costs associated with resolving a claim. In 2010, the increase in estimated pool severity was based on the resolution of defaults that occurred in prior periods with higher claim amounts, which in part, were applied to remaining deductibles on certain pool policies.

The "Losses paid" section of the table above shows the breakdown between claims paid on default notices received in the current year, claims paid on default notices received in prior years and the decrease in losses paid related to terminated reinsurance agreements as noted in footnote (6) of the table above. It has historically taken, prior to the last few years, on average, approximately twelve months for a default which is not cured to develop into a paid claim, therefore, most losses paid relate to default notices received in prior years. Due to a combination of reasons that had slowed the rate at which claims are received and paid, including foreclosure moratoriums and suspensions, servicing delays, court delays, loan modifications, our fraud investigations and our claim rescissions and denials for misrepresentation, it is difficult to estimate how long it may take for current and future defaults that do not cure to develop into paid claims. In 2012 and 2011, we experienced an increase in claims paid on default notices received in the current year due to fewer claim investigations and an increase in short sales.

The liability associated with our estimate of premiums to be refunded on expected claim payments is accrued for separately at December 31, 2012 and 2011 and approximated $134 million and $114 million, respectively. Separate components of this liability are included in "Other liabilities" and "Premium deficiency reserve" on our consolidated balance sheet. (See Note 3 – "Summary of Significant Accounting Policies – Revenue Recognition")

A rollforward of our primary default inventory for the years ended December 31, 2012, 2011 and 2010 appears in the table below. The information concerning new notices and cures is compiled from monthly reports received from loan servicers. The level of new notice and cure activity reported in a particular month can be influenced by, among other things, the date on which a servicer generates its report, the number of business days in a month and by transfers of servicing between loan servicers.


	2012			2011			2010


Default inventory at beginning of period		175,639			214,724			250,440
New Notices		133,232			169,305			205,069
Cures		(120,248	)		(149,643	)		(183,017	)
Paids (including those charged to a deductible or captive)		(45,741	)		(51,138	)		(43,826	)
Rescissions and denials		(3,037	)		(7,609	)		(13,942	)
Default inventory at end of period		139,845			175,639			214,724

Pool insurance notice inventory decreased from 32,971 at December 31, 2011 to 8,594 at December 31, 2012. During the third quarter of 2012, approximately 15,600 pool notices were removed from the pool notice inventory due to the exhaustion of the aggregate loss on a pool policy we have with Freddie Mac. See Note 20 – "Litigation and Contingencies" for a discussion of our settlement with Freddie Mac regarding this pool policy. The pool insurance notice inventory was 43,329 at December 31, 2010.

The decrease in the primary default inventory experienced during 2012 and 2011 was generally across all markets and all book years. However the percentage of loans in the inventory that have been in default for 12 or more consecutive months has increased, as shown in the table below. Historically as a default ages it becomes more likely to result in a claim.

Aging of the Primary Default Inventory

	December 31,				December 31,				December 31,
	2012				2011				2010

Consecutive months in default
3 months or less		23,282	17	%		31,456	18	%		37,640	18	%
4 - 11 months		34,688	25	%		46,352	26	%		58,701	27	%
12 months or more		81,875	58	%		97,831	56	%		118,383	55	%

Total primary default inventory		139,845	100	%		175,639	100	%		214,724	100	%

Primary claims received inventory included in ending default inventory (1)		11,731	8	%		12,610	7	%		20,898	10	%

(1) Our claims received inventory includes suspended rescissions as discussed in Note 20 – "Litigation and Contingencies". In connection with the Countrywide proceedings, we have voluntarily suspended rescissions of coverage related to loans that we believe could be included in a potential resolution. As of December 31, 2012, coverage on approximately 2,150 loans, representing total potential claim payments of approximately $160 million, that we had determined was rescindable were affected by our decision to suspend such rescissions. Substantially all of these potential rescissions relate to claims received beginning in the first quarter of 2011 or later. As of December 31, 2012, coverage on approximately 250 loans, representing total potential claim payments of approximately $17 million, were affected by our decision to suspend rescissions for another customer for which we also consider settlement probable. In addition, as of December 31, 2012, approximately 240 rescissions, representing total potential claim payments of approximately $16 million, were affected by our decision to suspend rescissions for customers other than those for which we consider settlement probable, as defined in ASC 450-20.

The length of time a loan is in the default inventory can differ from the number of payments that the borrower has not made or is considered delinquent. These differences typically result from a borrower making monthly payments that do not result in the loan becoming fully current. The number of payments that a borrower is delinquent is shown in the table below.

Number of Payments Delinquent

	December 31,				December 31,				December 31,
	2012				2011				2010


3 payments or less		34,245	24	%		42,804	24	%		51,003	24	%
4 - 11 payments		34,458	25	%		47,864	27	%		65,797	31	%
12 payments or more		71,142	51	%		84,971	49	%		97,924	45	%

Total primary default inventory		139,845	100	%		175,639	100	%		214,724	100	%

Rescissions

Before paying a claim, we can review the loan file to determine whether we are required, under the applicable insurance policy, to pay the claim or whether we are entitled to reduce the amount of the claim. For example, all of our insurance policies provide that we can reduce or deny a claim if the servicer did not comply with its obligation to mitigate our loss by performing reasonable loss mitigation efforts or diligently pursuing a foreclosure or bankruptcy relief in a timely manner. We also do not cover losses resulting from property damage that has not been repaired.

In addition, subject to rescission caps in certain of our Wall Street bulk transactions, all of our insurance policies allow us to rescind coverage under certain circumstances. Because we can review the loan origination documents and information as part of our normal processing when a claim is submitted to us, rescissions occur on a loan by loan basis most often after we have received a claim. Prior to 2008, rescissions of coverage on loans were not a material portion of our claims resolved during a year. However, beginning in 2008, our rescissions of coverage on loans have materially mitigated our paid losses. In each of 2009 and 2010, rescissions mitigated our paid losses by approximately $1.2 billion; in 2011, rescissions mitigated our paid losses by approximately $0.6 billion; and in 2012, rescissions mitigated our paid losses by approximately $0.3 billion (in each case, the figure includes amounts that would have either resulted in a claim payment or been charged to a deductible under a bulk or pool policy, and may have been charged to a captive reinsurer). In recent quarters, less than 10% of claims received in a quarter have been resolved by rescissions, down from the peak of approximately 28% in the first half of 2009.

Our loss reserving methodology incorporates our estimates of future rescissions and reversals of rescissions. Historically, the number of rescissions that we have reversed has been immaterial. A variance between ultimate actual rescission and reversal rates and our estimates, as a result of the outcome of claims investigations, litigation, settlements or other factors, could materially affect our losses. We estimate rescissions mitigated our incurred losses by approximately $2.5 billion in 2009 and $0.2 billion in 2010. In 2011, we estimate that rescissions had no significant impact on our losses incurred. All of these figures include the benefit of claims not paid in the period as well as the impact of changes in our estimated expected rescission activity on our loss reserves in the period. In the fourth quarter of 2012, we estimate that our rescission benefit in loss reserves was reduced due to probable rescission settlement agreements and that other rescissions had no significant impact on our losses incurred in 2012. For information about two settlements that we believe are probable, as defined in ASC 450-20, see Note 20 – "Litigation and Contingencies." The completion of those settlements, assuming they occur, may encourage other lender-customers to seek remedies against us.

We do not utilize an explicit rescission rate in our reserving methodology, but rather our reserving methodology incorporates the effects rescission activity has had on our historical claim rate and claim severities. A variance between ultimate actual rescission rates and these estimates could materially affect our losses incurred. Our estimation process does not include a direct correlation between claim rates and severities to projected rescission activity or other economic conditions such as changes in unemployment rates, interest rates or housing values. Our experience is that analysis of that nature would not produce reliable results, as the change in one condition cannot be isolated to determine its sole effect on our ultimate paid losses as our ultimate paid losses are also influenced at the same time by other economic conditions. The estimation of the impact of rescissions on incurred losses must be considered together with the various other factors impacting incurred losses and not in isolation.

The table below represents our estimate of the impact rescissions have had on reducing our loss reserves, paid losses and losses incurred.

	2012			2011		2010
	(In billions)

Estimated rescission reduction - beginning reserve	$	0.7		$	1.3	$	2.1

Estimated rescission reduction - losses incurred (1)		(0.2	)		-		0.2

Rescission reduction - paid claims		0.3			0.6		1.2
Amounts that may have been applied to a deductible		-			-		(0.2	)
Net rescission reduction - paid claims		0.3			0.6		1.0

Estimated rescission reduction - ending reserve	$	0.2		$	0.7	$	1.3

(1) As discussed in Note 20 – "Litigation and Contingencies" we have made substantial progress in reaching an agreement with Countrywide to settle the dispute we have concerning rescissions. We have determined that a settlement with Countrywide is probable and under this probable settlement agreement, we estimate that our rescission benefit in loss reserves was reduced by approximately $0.2 billion. This adjustment to loss reserves was partially offset by an estimated benefit of approximately $0.1 billion under the terms of the settlement agreements we consider probable, as defined in ASC 450-20. As a result, in the fourth quarter of 2012, we increased our loss reserve estimates by approximately $0.1 billion to reflect potential settlements we consider probable.

At December 31, 2012, our loss reserves continued to be impacted by expected rescission activity. We expect that the reduction of our loss reserves due to rescissions will continue to decline.

If the insured disputes our right to rescind coverage, the outcome of the dispute ultimately would be determined by legal proceedings. Under our policies, legal proceedings disputing our right to rescind coverage may be brought up to three years after the lender has obtained title to the property (typically through a foreclosure) or the property was sold in a sale that we approved, whichever is applicable, although in a few jurisdictions there is a longer time to bring such an action. For the majority of our rescissions since the beginning of 2009 that are not subject to a settlement agreement, this period in which a dispute may be brought has not ended. Until a liability associated with a settlement agreement or litigation becomes probable and can be reasonably estimated, we consider a rescission resolved for financial reporting purposes even though legal proceedings have been initiated and are ongoing. Although it is reasonably possible that, when the proceedings are completed, there will be a determination that we were not entitled to rescind in all cases, we are sometimes unable to make a reasonable estimate or range of estimates of the potential liability. Under ASC 450-20, an estimated loss from such proceedings is accrued for only if we determine that the loss is probable and can be reasonably estimated. Therefore, when establishing our loss reserves, we do not generally include additional loss reserves that would reflect an adverse outcome from ongoing legal proceedings. For more information about these legal proceedings, see Note 20 – "Litigation and Contingencies."

The liability associated with our estimate of premiums to be refunded on expected future rescissions is accrued for separately. At December 31, 2012 and 2011 the estimate of this liability totaled $18 million and $58 million, respectively. Separate components of this liability are included in "Other liabilities" and "Premium deficiency reserve" on our consolidated balance sheet. Changes in the liability affect premiums written and earned and change in premium deficiency reserve.

In April 2011, Freddie Mac advised its servicers that they must obtain its prior approval for rescission settlements and Fannie Mae advised its servicers that they are prohibited from entering into such settlements. In addition, in April 2011, Fannie Mae notified us that we must obtain its prior approval to enter into certain settlements. Since those announcements, the GSEs have approved our settlement agreement with one customer and have rejected settlement agreements that were structured differently. We have reached and implemented settlement agreements that do not require GSE approval, but they have not been material in the aggregate.

As discussed in Note 20 – "Litigation and Contingencies" we have made substantial progress in reaching an agreement with Countrywide concerning rescissions. In addition to the proceedings involving Countrywide, we are involved in legal proceedings with respect to rescissions that we do not consider to be collectively material in amount. We continue to discuss with other customers their objections to material rescissions and have reached settlement terms with several of our significant customers. In connection with some of these settlement discussions, we have suspended rescissions related to loans that we believe could be included in potential settlements. As of December 31, 2012, approximately 240 rescissions, representing total potential claim payments of approximately $16 million, were affected by our decision to suspend rescissions for customers other than the two customers for which we consider a settlement agreement probable, as defined in ASC 450-20. Although it is reasonably possible that, when the discussions or legal proceedings with customers regarding rescissions are completed, there will be a conclusion or determination that we were not entitled to rescind in all cases, we are unable to make a reasonable estimate or range of estimates of the potential liability.