Scant Earnings Relief as Banks Build Reserves
Not all individual results will be so dire. On Wednesday, Oct. 14, JPMorgan Chase (JPM) will be the first big bank to report. Analysts expect Chase to turn in earnings of 52¢ a share, compared with a loss of 6¢ a year ago, according to Thomson Reuters. Bank of America (BAC), on the other hand, is expected to report a loss on Oct. 16 of 4¢ a share, vs. a profit of 15¢ in the same quarter last year, according to Thomson Reuters (TRI).
The industry's woes particularly stem from its need to build reserves for souring loans of all types—to consumers, businesses, and commercial real estate owners. Bank executives have already warned investors that they spent more for such provisions during the quarter. The bankers are reacting to increasing delinquencies and to the fact that more troubled loans have moved beyond rehabilitation. KBW expects the industry's median cost for the provisions to come in at about double last year's figure.
the losses were well-anticipated The amount of provisioning expenses will vary a lot by bank, of course. BofA will likely report setting aside about $11.2 billion in the third quarter for loan losses, nearly twice as much as in last year's third quarter, estimates analyst Chris Kotowski of Oppenheimer (OPY). JPMorgan Chase is likely to up its expenses, too, but not by as much, setting aside $7.9 billion, compared with $5.8 billion last year.
Many analysts say this quarter's results have been so well-anticipated that what's most likely to move stock prices during the reporting season is what bank executives say about the loss provisions they'll be taking in the fourth quarter. Wall Street hopes bankers will drop hints that they're about to have the recession's version of a "blow-out quarter"—one in which they clear out the losses, as opposed to stunning the Street with amazing profits.
"I think they're going to be very aggressive in the fourth quarter about recognizing problems and reserving for them so they won't have that drag on earnings" in 2010, says Kotowski of Oppenheimer.
The Securities & Exchange Commission frowns on decisions that shift losses from one quarter to another, but bankers have a fair amount of discretion when recording costs for bad loans. They can estimate how much is reasonable to set aside in advance for loans expected to go bad. Later they can decide when to give up on a specific loan and how much of it to count as a loss to reserves. (It helps to have facts on hand to support their estimates and judgments in case auditors and regulators check their numbers.)
consumer loans may be the best betAdditional facts will be made clear in the coming weeks, and Kotowski believes bankers should offer a pretty good read later this year on the ultimate damage. Given how quickly bank assets are souring now, he says, "the final rightsizing of reserves" will likely take place in the fourth quarter. That would be the 11th-straight quarter banks have built reserves in spite of rising losses. By then, Kotowski figures, reserves at the five biggest banks in the U.S. will be up to 3.4% of their outstanding loans—about twice the norm.
Is this wishful thinking, like a food-poisoning victim hoping that each heave will be the last? A predictable end to losses is most plausible in the consumer-loan category. If unemployment stops rising later this year and if house prices hold steady—as many economists expect—it's fair to think bankers will then have the right numbers for loans on credit cards, cars, and houses.
The biggest doubts revolve around the banks' loans on apartment complexes, shopping centers, and office buildings. Commercial real estate losses are always the last to sink to lows after a recession because multiyear tenant leases delay the inevitable declines in building cash flows. Some real estate experts say it could be at least another year before building prices find their lows. That means it could take that long before bankers can reliably estimate how much they'll lose in foreclosures—and whether the reserves they will take next quarter are enough.