In November 2010, we estimated the total taxpayer cost to keep the GSEs solvent at about $280 billion, including $207 billion of credit losses embedded in the combined portfolios of Fannie and Freddie, as of June 30, 2010. Our projection for the ultimate cost to resolve Fannie Mae and Freddie Mac hasn’t changed. But we’ve updated our loss assumptions to reflect the housing market’s performance since then plus GSE-specific data through March 31, 2011. Based on recent data, we now believe there to be $190 billion of lifetime losses remaining in Fannie and Freddie’s combined single-family portfolios. When we add the $22.3 billion in combined credit losses Fannie and Freddie have taken since our last estimate on June 30, 2010 data, to our updated estimate, our total credit loss expectation tracks very closely with our previous estimate.
Fannie Mae and Freddie Mac have supported most mortgage funding in the U.S. since the domestic housing crisis arrived in 2007. Combined with Ginnie Mae (not rated), the GSEs have guaranteed more than 90% of single-family conforming mortgages since 2007. They benefit from the U.S. government’s effective guarantee of their obligations because of their public policy role, and have relied on capital injections from the U.S. Treasury as their losses mounted, eroding their capital. The U.S. Treasury has said it will stand behind Fannie’s and Freddie’s obligations, and will continue to provide them with necessary capital, but has also voiced its desire to downsize and re-focus these entities. Until then, these entities must continue to manage existing exposures lower while providing liquidity to the market in order to minimize future taxpayer costs.
The combined balance sheets of Fannie and Freddie support more than half of all existing mortgages in the U.S. At a time when home prices continue to fall, unemployment rates remain high, and economic growth is sluggish, the GSEs remain far from recovery, in our view.
Legacy Assets Continue To Hinder Earnings
The performance of Fannie and Freddie diverged somewhat in the first quarter of 2011. Fannie reported a net loss of $6.5 billion (before recording $2.2 billion in preferred dividends), while Freddie reported a net profit of $676 million (before recording $1.6 billion in preferred dividends).
At the same time, Freddie’s capital position benefitted from a $2.1 billion improvement in accumulated other comprehensive income (AOCI), reflecting gains on its available for sale (AFS) securities. It ended the quarter with a net worth of $1.2 billion, and did not need to request additional funding from the U.S. Treasury.
Fannie, on the other hand, recorded only a $181 million improvement in AOCI and ended the period with a net worth deficit of $8.4 billion. The acting director of the Federal Housing Finance Authority (FHFA) requested an additional $8.5 billion in capital for Fannie from the Treasury. Therefore, we expect Fannie’s obligation to the Treasury to rise to $99.7 billion, while Freddie’s obligation will remain $64.7 billion. Combined, the U.S. Treasury had invested $164.4 billion in the two through March since taking the GSEs into conservatorship in September 2008.
Both firms benefitted from slower mortgage prepayments as mortgage rates rose in the quarter, improving net interest income. At the same time, their funding costs remain low as they can still continuously refinance higher-rate long-term debt. Both companies also continue to purchase seriously delinquent mortgages out of their securitized pools because the cost of funding these purchases is much lower than the interest payments to investors that would be required on outstanding non-accruing loans. Lower returns in each company’s investment portfolios, reflecting lower interest income, are partially offsetting those gains.
The inventory of real estate owned (REO) properties at Freddie Mac declined 9.6% to 65,174 units (worth $6.3 billion) as of March 31, 2011, from 72,093 units on Dec. 31, 2010, because of a pick up in sales of foreclosed properties during the first quarter. This was true at Fannie Mae also, where the inventory declined 5.9% to 153,224 units (worth $14.1 billion) from 162,489 units for the same periods. However, the GSEs could take on more properties this year, given that mortgage servicers have resumed foreclosure activities after a temporary hiatus because of various bank documentation and process irregularities. Loss severity ratios (which still measure in the high 30% area) on properties sold have risen slightly, reflecting weak housing prices. They could continue to rise as foreclosed properties stay on the market longer.
The Credit Quality Of Single-Family Mortgage Loan Books Is Gradually Improving
The credit quality of portfolios at both institutions has improved. Fewer loans are newly past-due, which has reduced the number of loans at risk of becoming seriously delinquent. The SDQ rate in Freddie Mac’s single-family mortgage portfolio dropped to 3.63% in the first quarter of 2011 from 3.84% in the fourth quarter of 2010, and to 4.27% from 4.48% at Fannie Mae. Both GSEs have experienced sequential quarterly improvement in SDQ rates since the first quarter of 2010, when they were 4.13% for Freddie Mac and 5.47% for Fannie Mae, except for a slight blip up at Freddie in the fourth quarter of 2010. We attribute the improvement to their efforts to help overextended borrowers hold onto their homes with mortgage modifications and other foreclosure alternatives. Moreover, the resumption of foreclosures has helped clear SDQ loans. However, Fannie Mae still reports that 71% of its SDQ loans have been delinquent more than 180 days (a year ago that figure was 62%). This indicates that there’s still a significant pipeline to process through modification or foreclosure. (Freddie Mac does not report a comparable figure.)
Weak Economic Conditions Are The Main Factor Behind Serious Delinquencies
The number of seriously delinquent loans will likely continue to fluctuate with home prices, unemployment rates, and household wealth, and with the extent to which borrowers with modified loans become delinquent again. SDQ rates will also remain higher than they would be otherwise because the GSEs require servicers to pursue home retention and modification alternatives for qualifying borrowers before moving to foreclosure. Moreover, the length of time for which loans are remaining seriously delinquent continues to increase, in part because of foreclosure delays.
There are caveats to our estimates. We don’t take into account sensitivity to home price indices (HPI), though we expect the recent double dip in home prices, reflected in the S&P/Case-Shiller Home Price Indices, to increase mortgage delinquencies and roll rates to SDQ over the next six months. Still, we believe our estimates are sufficiently conservative. We assume that embedded losses in the GSEs’ loan portfolios are for the life of the portfolio--for mortgage assets that have an average life of 7-10 years. However, we don’t estimate when the GSEs will recognize losses. That’s because GSE servicers try to limit foreclosures, and because recent mortgage loan modifications have proved more successful than earlier ones, which we think reflects improved processes. For instance, more than 68% of loans modified in the first quarter of 2010 were performing nine months after modification, compared with about 34% of loans modified in the first quarter of 2009.
We expect to see marginally higher loss severity rates on foreclosures through 2011, because of the continued fragility in the housing market and general economy. A measured pace of foreclosures, however, could mitigate the effects of a higher rate of foreclosures. Foreclosure starts declined significantly late in 2010, as banks and servicers dealt with process and documentation issues. But some easing of that gridlock has allowed foreclosures to proceed and banks’ OREO (other real estate owned) to flow back into the market at a measured pace. We expect that renewed short sale and distressed purchase activity may help shorten foreclosure timelines, which should help to reduce the number of SDQ loans and provide more clarity on the near-term path of the housing market.
Because mortgage rates have remained historically low, refinancings have been the lion’s share of mortgage originations. The most common is still the 30-year fixed-rate conventional conforming loan that originators sell to the GSEs to limit their own interest rate risk. The performance and profitability of mortgages the GSEs have acquired since 2009 remain superior to those of previous vintages over corresponding seasoning periods. In our opinion, this primarily reflects more selective acquisition and more conservative underwriting, resulting in higher average borrower FICO scores for borrowers and lower average loan to value (LTVs) ratios--about five percentage points lower. A positive for the GSEs: Mortgages originated since 2009 now comprise 43% of their combined $4.6 trillion single-family mortgage portfolio. By contrast, nearly two-thirds of the credit losses have come from the 2007 and 2006 vintages, which now represent only about 19% of the combined single-family book of loans. But we believe these earlier vintages will continue to account for about one third of the remaining loss embedded in the portfolio of performing loans.
The Economic Recovery Is Still Modest At Best
Although we believe that loan performance appears to be stabilizing, as evidenced by fewer delinquencies, the biggest risk could be persistently high unemployment or stalling economic growth. Our economists estimate that the weak second-quarter GDP reading has reduced growth for 2011 to 2.6%, barely above our long-term trend growth rate of 2.5%--and not enough to make a dent in unemployment. The economy needs growth of 1% above trend to bring the unemployment rate down by 0.5%. Our economists expect unemployment to remain above 8% into 2013.
The weaker GDP growth, job creation, and housing data in April and May are also dampening consumer confidence and that could lead to more strategic defaults (in which borrowers walk away from homes whose mortgages exceed their market value). Strategic defaults could also resume if home prices or price appreciation decline. Currently, about one in four mortgages is higher than the value of the home. The Case/Shiller Index is now 33.1% below its July 2006 peak. It is also below the previous nadir of 32.6% set in April 2009, officially confirming a double dip in U.S. home prices. Twelve cities have posted new record lows. We are seeing the double dip stress the housing market further, especially in those regions with higher foreclosure rates and general economic weakness (e.g. Michigan, Arizona, Illinois, and Georgia). Standard & Poor’s expects prices to drop another 4% on average by year end. We also expect the FHFA home price index to drop another 5% and not to hit bottom until early 2012. We believe the GSEs have incorporated similar assumptions in their reserving actions, which has led to higher credit-related costs early in 2011.
The Bottom Line
Profitability of more recent loans that the GSEs acquire (43% of their portfolios) will be at least break-even in our opinion. But we also recognize that the companies’ GAAP results could shift as interest rates rise because of the impact of derivative valuations on the GSEs’ earnings. We also believe that market movements trending higher will improve mark-to-market securities gains, offset by deeper credit-related impairments in the GSEs’ private label mortgage portfolios. We continue to believe that neither housing GSE has enough earnings to keep paying its dividend to the U.S. Treasury under their senior preferred stock purchase agreements. We thus expect that the FHFA will continue to ask the Treasury for capital draws on their behalf in amounts approximating their quarterly (and rising) dividend. This could amount to $2.2 billion for Fannie Mae and $1.6 billion for Freddie Mac--plus or minus quarterly derivative results and securities gains and losses. Each company’s Treasury stock purchase agreement offers unlimited funding through 2012. However, it limits the government’s funding commitment after 2012 to $200 billion less certain previous draws and positive net worth.
The Taxpayer Cost Could Be Slightly Lower Than Our Estimate
We believe the total cost to resolve Fannie Mae and Freddie Mac is tracking closely with our initial estimate of $280 billion. Of course, smaller projected losses on the GSEs’ mortgage portfolios, as well as smaller lifetime losses from their securities portfolios and counterparty exposures could lower this amount. To be sure, the cost to taxpayers to resolve these entities remains sizable. But we think the GSEs have made some progress toward limiting the damage.
"
No comments:
Post a Comment