Bank Asset Quality Update

Posted by jeff

Tue Aug 19th, 2008 08:41 PM

The data on bank asset quality are out from the Federal Reserve Bank of San Francisco and are presented below in graphic form going back to the last real estate debacle in the early 90s. While the data are from federally chartered savings banks and are not for all banks, it is a large enough sample size to be representative of what is happening with banks overall. Please note that savings & loans and credit unions have traditionally had a much larger percentage of residential mortgage exposure and a much smaller percentage of commercial and industrial loans and commercial real estate loans. Recall that while these data points are lagging because most banks that are publicly traded have reported results for the quarter, including some version of the types of numbers reported here; we consider this information to be very important in this particular crisis. This is because so many banks were involved in the CMBS market, wherein they packaged loans for sale and these loans or pieces of these loans appear on bank balance sheets as assets held for sale or trading assets and must be accounted for on a mark-to-market basis. This means that their values have been impaired by trading action in the markets related to assumptions regarding losses, but they don't really tell the story of whether the actual borrowers are still current on the loans underlying the securities. The only way to get data on this is to go through the static pool data that is filed on securitizations one at a time to get a picture of what is going on. Additionally, these securitized assets tended to be the most poorly underwritten. In contrast, the data below relate to loans that are of the highest quality; after all, these are the loans banks chose to hold on their books, for better or for worse, until they mature or default. These data also give us a broad cross section of data points from across the country, and we have great historical information to compare it to. If you're a data nerd like me....this is exciting stuff, if not you may want to move on to the latest Olympics news. Please excuse the large charts but because it is over 10 years of data it is very hard to see the scales at lower resolutions.


Okay. Residential delinquencies are an outright disaster, even the paper banks chose to hold themselves and not package up for sale to unsuspecting bond and hedge funds. Not that surprising, but the trajectory is like the space shuttle taking off on a clear day. No slowdown in the rate at which this paper is going bad is in sight and we are now well above the levels hit in the early 90s.


Residential charge-offs tend to lag delinquencies by a quarter or two. It's no surprise that these have gone ballistic recently. Note that charge-offs are sometimes reversed at a later date depending on how much value the bank gets when it re-sells the underlying real estate that backed the loan. The "severtity" of its loss will vary based on what the loan to value was on the loan and how badly real estate prices have declined.


Credit card delinquencies actually declined slightly quarter-to-quarter. It is possible that this is seasonal, as people tend to have shopping hangovers from Christmas in q1 and delinquencies may settle back some in Q2. I take any improvement in consumer debt delinquencies to be a positive in this environment. Remember that credit card receivables tend to be outstanding for a shorter time than mortgages and credit card companies are aggressively managing how much credit they are extending to customers and what punitive rates and fees they impose to try to keep consumers from self destructing.


Credit card charge offs played a little catch up this quarter. My guess is that surging bankruptcies associated with real estate losses may be sending some customers straight to the charge-off counter, leaving no skid marks at the delinquency desk.


Perhaps most important are the commercial real estate statistics. This category is what practically brought down the financial system in the early 90s, as you can see from the mountain of problem loans evident in the early 90s period on the chart above. Thus far it has held up much better than the residential area. But delinquencies broke out of a long slumber in Q4 and surged in Q1. The Q2 commercial delinquency figure is up a little but not surging like it's residential cousin; this is very good news indeed. Delinquencies are still far below the dark days of the early 1990s.


To some degree I can only speculate on the huge surge in commercial real estate charge offs. It takes some time for delinquencies to move to the charge off bucket and there is probably some cumulative effect from this lag. Additionally it is possible that a couple of large project failures moved quickly from delinquency and into the charge-off bucket. Without a further big increase in delinquencies, I would actually expect this number to drop back some next quarter.


At the end of the day, the total number of delinquent loans as a percent of all loans is the most important figure in assessing risks to bank solvency. While this number is moving up, it grew .27 basis points in Q2 2008 to 3.16%; this is a slower increase than the last 3 quarters, which ranged from .33 basis points to .43 basis points since beginning to climb in Q207. Note however, that at 3.16% of all loans, this is the highest level of delinquencies registered since the 3.25% level hit in Q1 of 1994. At that time, the economic/bank recovery was driving delinquencies below the 4% plus level sustained during the S&L crisis and commercial real estate crisis period stretching from 1985 to 1993. My key takeaways are that we have not reached these crisis levels yet, though it seems like we have a fair shot of getting there. Notice that we are back nearing the levels seen at the depths of the dot com bust, despite the lay off picture being less severe thus far in the downturn. Additionally, it is important to note that these horrible levels of non-performing loans were sustained by the banking system for 8 years during the last real estate down cycle and we didn't fall into a depression. So maybe there is still some hope.


All loan charge offs as a percentage of all loans have surged. But like the commercial real estate case it is probably a function of lumpiness and the lag versus delinquencies. I think banks are just starting to get more serious about moving delinquent loans into this category.

All in all the data are more of the same thing we have been seeing. The surge in residential delinquencies is alarming, while the more sedate increase in commercial real estate delinquencies and the decline in credit card delinquencies is slightly reassuring with regard to the spread of the credit pandemic to other asset classes. I will start to really worry if we see a surge in commercial real estate delinquencies, and especially if overall loan delinquencies get back to the early 1990s levels. Take small solace in the fact that we are not quite there yet.