by Anonymous on 12/11/2007 11:08:00 AM

There have been some questions about what this means (thanks, Ramsey, for bringing it to my attention):

MCLEAN, Va., Dec. 10 /PRNewswire-FirstCall/ -- Freddie Mac (NYSE: FRE - News) announced today that the company will generally purchase mortgages that are 120 days or more delinquent from pools underlying Mortgage Participation Certificates ("PCs") when:

-- the mortgages have been modified;
-- a foreclosure sale occurs;
-- the mortgages are delinquent for 24 months;
or
-- the cost of guarantee payments to security holders, including advances
of interest at the security coupon rate, exceeds the cost of holding
the nonperforming loans in its mortgage portfolio.

Freddie Mac had generally purchased mortgages from PC pools shortly after they reach 120 days delinquency. From time to time, the company reevaluates its delinquent loan purchase practices and alters them if circumstances warrant.

Freddie Mac believes that the historical practice of purchasing loans from PC pools at 120 days does not reflect the pattern of recovery for most delinquent loans, which more often cure or prepay rather than result in foreclosure. Allowing the loans to remain in PC pools will provide a presentation of its financial results that better reflects Freddie Mac's expectations for future credit losses. Taking this action will also have the effect of reducing the company's capital costs. The expected reduction in capital costs will be partially offset by, but is expected to outweigh, greater expenses associated with delinquent loans.

Attentive readers Calculated Risk addicts will remember the big tizzy last month over Fannie Mae's presentation of its credit loss ratio, and the confusion-party in the press that it engendered. As you recall, the issue was that Fannie Mae had been exercising its option (not obligation) to buy delinquent loans out of its MBS in order to pursue workout efforts with them. Because that requires the MBS to be paid off at par, but the accounting rules require Fannie to take these loans to their portfolio at the lower of cost (par) or fair market value, and since "fair market value" right now for a delinquent loan, even one you think can be cured (made reperforming with a workout), is terrible, doing this means that Fannie booked big write-downs at the time. Fannie then backed some, but not all, of that write-down out of its credit loss ratio on the portfolio, to distinguish between true foreclosure-related charge-offs and these fair value adjustments, and uproar ensued, with Fortune's Peter Eavis implying that they were cooking the books and hiding losses.

It appears to me that Freddie Mac has decided it doesn't want to go there. It is therefore doing what, presumably, Peter Eavis wants it to do: leave the delinquent loans in the MBS pools unless and until that becomes more expensive than taking them out.

The accounting here is rather complex (which won't stop some people from having a cow over it, but I can't help that). The somewhat simplified view is this: the GSEs guarantee timely payment of principal and interest to MBS investors. They do not guarantee that investors will earn interest forever, but only as long as principal is invested. If a loan payment is not made by the borrower, either the GSE or the servicer (depending on the contract) has to advance scheduled principal and interest payments to the MBS until such time as the loan catches up (the borrower makes up the past due payments) or is foreclosed and liquidated.

The GSEs collect guarantee fees from seller/servicers (it works like servicing fees: it comes off the monthly interest payment). They also collect some other lump-sum fees when pools are settled. This is revenue to the GSEs, with a corresponding liability (to make the advances, with the risk that the advances will not be reimbursed completely at liquidation of the loan).

Therefore, when there are deliquent loans in an MBS, and the servicer is not obligated to advance for them, the GSE has a choice: it can buy the loan out of the MBS, put it into the GSE's own portfolio, and take any and all losses directly as any other investor would (and also any income). Or, it can leave the loan in the pool, while advancing the scheduled P&I to the pool investors. Only in some specific cases is the GSE obligated to take out a loan: when mortgages are modified, when the foreclosure sale occurs, or when the loan has been delinquent for 24 months or more. In other situations, it comes down to the question of which is cost-effective for the GSE: to leave it there and continue to advance, or to take it out with portfolio capital and do the fair value write-down.

Do note that if Fannie Mae had adopted this policy that Freddie has just announced--basically, that buying the loan out of the MBS will be the last rather than the first resort--it would not have shown big fair value write-downs, those would not have affected the credit loss ratio, and a big dust-up would not have occurred. There would still have been an effect on the financials, but just in a different place: in advances (coming out of G-fees received). So you either have expense (P&I advanced to bondholders) or you have expense (capital used to buy out the loans).

Insofar as everyone has been all worked up about the GSEs' capital ratios, this should be good news: they are levering investors' capital to carry delinquent loans until they can be cured (or liquidated). Insofar as investors want principal back faster, it's maybe not good news. But you can't really have it both ways.

I think it is important to understand that the GSEs are supposed to cover guarantee costs out of g-fees, not with portfolio purchases. You might have noticed that both Fannie and Freddie are increasing the g-fees and postsettlement/loan level pricing adjustments they charge seller/servicers. So they are beefing up the funds they have to cover MBS losses with. Nothing guarantees that will be enough; I don't think anybody knows right now what will be enough. But for what it's worth, I don't see this as "playing games" with capital requirements. I guess we'll have to see what Fortune Magazine thinks.

by Calculated Risk on 12/10/2007 09:29:00 PM

From the WSJ: Investor Withdrawals Shut BofA Fund

Columbia Management is shutting its Strategic Cash Portfolio ...

Only some investors will get their cash out. The fund's biggest investors will be paid "in kind" -- that is, they will be given their share of the underlying securities, rather than a cash payment. Smaller shareholders can cash out at the fund's share price, which is currently 99.4 cents on the dollar. The fund required a minimum investment of $25 million.

So small investors - with just $25 million - will receive 99.4 cents on the dollar. Large investors will be paid "in kind". Ouch!

Some enhanced funds are doing even worse:

A report Monday by Standard & Poor's found that about 30 U.S.-oriented enhanced cash funds rated by S&P had lost a total of $20 billion, or 25% of their assets, in the third quarter. In one of the more dramatic instances, one fund (which S&P declined to identify) saw its assets under management shrink by 98% or $2.5 billion.

Even traditional money-market funds have felt pressure. ... at least a half-dozen financial institutions, including Bank of America, have taken steps such as buying the funds' troubled securities to protect their money funds....

Money-market funds are required to maintain an unchanging $1-per-share net asset value; if they waver from that they are said to "break the buck." Enhanced cash funds don't have the same requirement.

by Calculated Risk on 12/10/2007 06:02:00 PM

Update: Here is the Morgan Stanley piece. (hat tip Carlomagno)

From Rex Nutting at MarketWatch: Mild recession likely, Morgan Stanley says

The U.S. economy is likely to slip into a mild recession in 2008, said economists at Morgan Stanley, which is the first major Wall Street firm to predict a recession.

Domestic demand is expected to fall 1% annualized over the next three quarters with zero growth in gross domestic product and a 5% to 10% drop in corporate earnings, said chief economist Richard Berner and U.S. economist David Greenlaw in an updated forecast on the firm's Global Economic Forum Web site. For the full year, Morgan Stanley sees 1% growth.
...
"Those negatives sound like the recipe for a serious recession, so why do we think it will be mild?" Berner and Greenlaw wrote. "Although it is slowing, global growth is still strong, and we expect that net exports will add about 3/4 percentage point to growth through the end of 2008. In addition, we think that corporate capital and hiring discipline in this expansion mean that there are no business-investment or labor-market excesses to unwind, adding to U.S. economic resilience."

Also moving into the recession camp, according to the Wall Street Journal, is Paul Kasriel of Northern Trust, the 2006 recipient of the Lawrence R. Klein Award for Blue Chip Forecasting Accuracy.

by Calculated Risk on 12/10/2007 04:59:00 PM

WaMu PR: WaMu to Raise $2.5 Billion in Additional Capital, Reduce Dividend, Resize Home Loans Business and Cut Expenses to Fortify Capital Base

Washington Mutual, Inc. (NYSE:WM - News) announced today a series of actions designed to address the unprecedented challenges in the mortgage and credit markets by strengthening the company’s capital and liquidity and accelerating the alignment of its Home Loans business with its retail banking operations.

These actions include:

A capital offering of convertible preferred stock with aggregate proceeds of approximately $2.5 billion;

A major reduction in company-wide noninterest expense of approximately $500 million for 2008 as a result of a substantial resizing of its Home Loans business and reduced corporate support expense; and

A significant change in the strategic focus of its Home Loans business in response to a changed market.

In addition, the company said its Board of Directors intends to reduce the quarterly dividend rate to $0.15 per share from its most recent quarterly dividend rate of $0.56 per share.
...
Continued deterioration in the mortgage markets and declining housing prices have led to increasing fourth quarter charge-offs and delinquencies in the company’s loan portfolio. As a result, the company now expects its fourth quarter provision for loan losses to be between $1.5 and $1.6 billion, approximately twice the level of expected fourth quarter net charge-offs.

The company currently expects its first quarter 2008 provision for loan losses to be in the range of $1.8 to $2.0 billion, reflecting an increase in provision well ahead of charge-offs, which are also expected to increase significantly during the quarter. The first quarter range reflects the company’s current view that prevailing adverse conditions in the credit and housing markets will persist through 2008.

While difficult to predict, the company also currently expects quarterly loan loss provisions through the end of 2008 to remain elevated, generally consistent with its expectation for the first quarter of 2008.

by Calculated Risk on 12/10/2007 01:46:00 PM

"Existing-home sales, after reaching the third highest total on record, 6.48 million in 2006, are forecast at 6.44 million in 2007 and 6.64 million next year."
National Association of Realtors, Feb 2007.

Just thought I'd start with a little comic relief from Those Wacky NAR Forecasts! Compare to the current NAR forecast:

Existing-home sales are likely to total 5.67 million this year, the fifth highest on record, rising to 5.70 million in 2008, in contrast with 6.48 million in 2006.
NAR, Dec 2007

I think their forecasting model is broken.

Last December it was hard to find a Wall Street firm with a forecast under 6.0 million existing home sales in 2007. David Berson, then Chief Economist at Fannie Mae, forecast 2007 sales would be 5.925 million - and that was considered bearish. My forecast was for sales to "surprise to the downside, perhaps in the 5.6 to 5.8 million unit range."

Now we are starting to see forecasts for 2008. Back in August, Goldman Sachs forecast existing home sales would fall to 4.9 million in 2008. However, since then, Goldman has becoming even more bearish on housing.

Here is another forecast for 2008 via AP:

Patrick Newport, an economist at Global Insight, forecasts that home sales will drop from 5.66 million this year to 4.7 million in 2008

I'll put together a forecast at the end of the year, but this is why I posted the chart of sales as a percent of owner occupied units this morning (see 2nd chart). At the bottom of a housing cycle, sales typically fall to 5% or even 4% of owner occupied units (a measure of turnover). There are currently just over 75 million owner occupied units in the U.S., so 6% (the median for the last 40 years) would be sales of about 4.5 million in 2008, 5% would be sales of 3.75 million, and 4% would be 3.0 million units. This isn't a forecast - just a review of historical data - but it is possible that sales could fall sharply from the current levels.