Alternative title: “How loan modifications distort bank results”.
Barclays Capital speculated last month that non-performing loans (NPLs) may peak in 2010. A few weeks on, and the bank is saying net-charge-offs (NCOs) may have peaked based on US banks’ recent fourth-quarter (2009) earnings. That was fast!
Here’s what BarCap analyst Jason Goldberg says (our emphasis):
In 4Q, the median bank’s net charge-off ratio IMPROVED 7bps from 3Q, better than our expectations of a 15bp increase (after losses came in higher than we expected in 3Q) and implying LOSSES MAY HAVE PEAKED IN 3Q09. This is particularly impressive since in 21 out of the prior 24 years, 4Q net charge-offs were the highest quarter of the year. Furthermore, the median bank’s NPA ratio increased 21bps linked quarter, the low end of our expectations of a 20-30bp increase, and much better than 52bp and 54bp increases in 3Q and 2Q, respectively. Also provisions declined 8%, modestly better than our expectation of a 5% decrease, and supporting our prior view that industry-wide provisions peaked in 2Q09. Still, we do not expect uniform improvement across our coverage and we look for consumer losses to peak before commercial. We are also concerned there will be a long tail to this real estate cycle.
But there’s a whole host of non-fundamental things that could be helping boost credit quality. For instance, there’ve been payment holidays on US credit cards and mortgages, distortions from the US Treasury’s Hamp programme, plus regulatory support for commercial real estate loan work-outs.
It’s a point not lost on Goldberg, who cautions:
Still, we wonder if recent asset metrics are not as good as advertised, particularly when comparing 4Q09 results to 3Q09. Investors need to be mindful that NPAs and NCOs: a) are at still extremely elevated levels; b) are still projected to increase at several banks in 1Q10 and NPA IN-FLOWS WERE POSITIVE FOR EVERY SINGLE BANK UNDER COVERAGE FOR THE 7TH STRAIGHT QUARTER IN 4Q09; and c) are potentially understated for some, amid loan modifications, payment holidays, restructurings, TDRs, extensions, foreclosure moratoriums and loan sales. While these moves might be the right course of action to take over-time, we believe the bigger impact will be to change the shape of the loss curve, more than the magnitude. As such, while the ultimate loss content may be moderately reduced, the time frame to get to normalized earnings (which continues to appear to be a focus of investors, see Figs 1&2), could be extended. It’s unclear to us if the market appreciates this phenomenon.
A good example of this is JPMorgan, according to Goldberg.
The bank reported NPAs declined 3 per cent and its NCO ratio improved 2bps in the fourth-quarter of 2009. Excluding loan sales, however, NPAs rose, and excluding the impact of a payment holiday in JPM’s credit card division, the overall NCO would have increased.
Back to the analyst:
. . . We use JPM as an example, not to pick-on them, but rather because we view that company as being relatively more transparent. We believe other banks are taking similar actions that are altering their credit quality metrics. Also note in November, regulators published a statement on CRE loan workouts making it easier to deal with CRE problem near-term as loss recognition based solely on a reduction in collateral value can be extended. It also promoted banks to restructure credits, rather than foreclose, similar to what we are witnessing on the residential front. With banks likely concentrating on past due loans, this could have the impact of making those metrics appear better than advertised. We also believe the 3Q09 share national credit exam elevated 3Q09 net charge-offs, making comps easier, while banks rushing to get TDRs done prior to year-end for beneficial accounting treatment, is another driver to watch. Note several banks exclude and/or don’t even report restructured loans, as well as TDRs, which we expect to have high re-default rates. Furthermore, others have been aggressively extending the maturity dates of certain residential real estate related loans, namely construction and lot loans, as well as using interest reserves. We also watching as foreclosures are expected to enter the market at an increasing pace. The combination of greater foreclosures and seasonal distortions could push home prices lower after surprisingly rising over the prior few months. Additionally, it appears a Second Lien Modification Program (2MP) is on the horizon.
Anyway, kicking the can of bad loans down the road isn’t really a huge problem, as long as redefault rates are relatively low and the (modified) loans are indeed paid off.
But evidence on that so far has been worrying. Some people have speculated, for instance, that the redefault rate for mortgages modified as part of Hamp could be as high as 60 per cent. The point is that it buys time by increasing the number of months between the borrower turning delinquent and the house coming onto the market. Such housing, owned by the banks and mortgage companies that foreclose, is called “shadow inventory” since it’s not included in official measures of unsold housing inventory.
So, Hamp-modified mortgages aside, are we now dealing with shadow loans on a wider scale?
Here are a couple of datapoints from Structured Finance News:
The redefault rate on modified commercial loans has been more than 50% in past cycles, and a federal report late last year showed that more than half of home mortgages modified in the third quarter of 2008 became delinquent within a year. Redefault rates vary from bank to bank. Fifth Third Bancorp, for instance, said in its fourth-quarter report that 25% of the $2.7 billion of loans it has modified since 2007 have gone bust.
Paul Miller, head of research at FBR Capital Markets Corp., said the poor success rate of modifications makes him worry that banks could be in for a second wave of nonperforming assets when their modified loans eventually sour.
Tom Mitchell, a senior analyst who covers financial stocks at Miller Tabak & Co., agreed that banks’ modified loans may be skewing the picture.
“We now have to consider the [modified loans] as a kind of shadow group of nonperforming assets,” Mitchell said. “It’s reasonable to say that for most banks, if the loans had not been [restructured], they would have been nonperformers.”
Note also, that many big banks have been lowering their provisions for loan losses.
Which means if a second wave of losses did come, they would not be as prepared as they could be.
(Full Barclays note in the Long Room)
Related links:
Loan modifications lurk as threat to credit quality gains – Structured Finance News
Hamp, what is it good for? – FT Alphaville
Great Depression-esque bad debt at banks – FT Alphaville
Regulators looking at banks’ coverage ratios, BarCap says