Fannie Mae's fuzzy math
The mortgage lender has quietly changed the way it calculates its bad loans -- and it could be camouflaging steep credit losses, writes Fortune's Peter Eavis.
(Fortune) -- Investors might want to take a closer look at Fannie Mae's latest earnings report. Lost in the unsurprising news of the mortgage lender's heavy losses was a critical change in the way the company discloses its bad loans -- a move that could mask that credit losses that are rising above levels that the company predicted just three months ago.
Fannie Mae's potentially misleading disclosure comes at a crucial time for the company. Fannie Mae was severely penalized last year for overstating earnings and for a lack of oversight. As part of its punishment, the amount of home loans that Fannie Mae can make was limited.
But now influential members of Congress, including Senator Charles Schumer, want Fannie Mae's watchdog, the Office of Federal Housing Enterprise Oversight (OFHEO), to temporarily lift the portfolio limits on the company and its rival Freddie Mac. Legislators want both lenders to buy more subprime mortgages to help stave off foreclosures.
Fannie Mae already holds a substantial amount of risky mortgages in its $2.4 trillion mortgage book -- and the recent shift in how it discloses a much-watched credit yardstick disguises just how quickly bad loans may be rising.
If that's the case, Fannie Mae will face a new barrage of questions about its bookkeeping.
Fannie Mae controller David Hisey responds that the change in how its loss numbers were presented makes them "more transparent, not misleading."
But Fannie Mae's numbers effectively make its credit look better than it is.
It all comes down to what's known as the credit loss ratio -- a measure that Fannie Mae has consistently provided to investors to help them assess the credit quality of its mortgages. The credit loss ratio expresses bad loan losses as a percentage of Fannie Mae's loans.
In August, Fannie Mae predicted its credit loss ratio would be 0.04-0.06 of a percentage point for all of 2007. (Wall Street generally refers to percentages in basis points, which each equal one hundredth of a percentage point. In Fannie Mae's terminology, then, its 2007 loss ratio estimate is four to six basis points.)
A range of four to six basis points may not sound like a big deal for an institution involved in mortgages, but for Fannie Mae it is the norm.
What matters is if Fannie Mae goes above that range. And Fannie Mae appears to have already done that this year. But its disclosure change makes that worrying development very hard to see.
Here's why: Last week, as part of its earnings report, Fannie Mae revealed that the company had changed the way it calculates the credit loss ratio. Under the new method, Fannie Mae's annualized credit loss ratio was just 4 basis points in the first nine months of the year.
At first glance, four basis points looks to be at the low-end of Fannie Mae's full-year forecast. Problem is, because the company is using a new methodology, the previous estimate no longer makes sense to use.
So what would have happened if the company had compared apples to apples -- and stuck with the old method of calculating its loss ratio?
Under the previous method, Fannie Mae would have been well outside of its range. The company would have reported an annualized loss ratio of 7.5 basis points in the first nine months of this year.
What exactly caused the change -- and how did it lead to a reduction in the credit loss ratio?
In its third quarter financial statements, Fannie Mae started to break out credit losses taken to fulfill an accounting treatment called SOP 03-3, which the company said it adopted at the start of 2005.
These SOP 03-3 losses were previously included in its credit loss ratio calculation, but Fannie Mae last week removed them from that calculation, causing its loss ratio to look much lower.
The company said it made the change to add transparency to its loss numbers and to show a more cash-based reflection of credit losses.
In a statement, Fannie Mae spokesman Brian Faith said that the forecast of four to six basis points was "predicated on our estimation of what our realized losses would be for the year."
What does realized losses mean? When asked that, Faith referred to Fannie Mae's most recent quarterly filing. There, realized losses appear to be defined as losses calculated under the new method. In other words, Faith appears to be suggesting that the 2007 forecast was always based on the new method of calculation.
Why is that hard to believe? When the company issued that range in August, it expressed all its published credit loss ratios under the old method. And on an August conference call, when discussing the full-year range of four to six basis points, Fannie Mae executives did not mention any change in calculation of the ratio.
Management acknowledges that credit losses are mounting. During an analyst call last week, Fannie Mae CEO Daniel Mudd warned that the company's loss ratio could rise to eight to 10 basis points in 2008, due to a worsening housing market. It's not clear whether that forecast is based on the old or new methodology.
The company may already be exceeding that 2008 guidance. Based on the old methodology for calculating the loss ratio for the third-quarter alone, the company's annualized loss ratio is already at 14 basis points.
If so, Fannie Mae's mounting losses are disturbing.
So what could a soaring loss ratio mean for Fannie Mae? Consider these numbers: At Sept. 30, Fannie Mae had exposure to $74 billion of loans with a FICO credit score below 620. Loans scored below 620 are generally classified as subprime. In addition, Fannie Mae has exposure to $196 billion of Alt-A mortgages, home loans for which the borrower doesn't have to submit complete documentation for basic criteria like income.
At the same time, Fannie Mae has only $40 billion of capital.