Wednesday, May 6, 2009

Preliminary Q1 2009 Stress Test Results: Significant Increase in Stress Across the Banking Industry

UPDATED: May 6, 2009

“Young man sometimes the only way to win is to lose gracefully.”

Gen. Bill Creech, Commander,
USAF Tactical Air Command

to Dennis Santiago years ago

Technology Innovation That Actually Works

We admire the Federal Deposit Insurance Corporation for many reasons. This month, we applaud them once again for the fine job they have done to bring the Central Data Repository (CDR) online in a way that serves the growing public demand for timely US bank data. Made operational just in January of 2009, we’ve just collected Q1 2009 CALL reports for over 7,600 reporting units, a goodly portion of which came out of submittal confirmation only a couple of days ago.

These CALL reports are submitted to the FDIC during a 30 day submission window beginning the first day of the quarter. We’ve been testing a variation of IRA’s ratings analyzer on these Q1 CALL’s since April 1st. The result is that IRA is now positioned to deliver preliminary bank stress estimates for the new quarter roughly two to three weeks ahead of the FDIC’s mid-quarter research master file release. Based on this maiden run, we are pleased to report that the FDIC’s CDR engine has the potential to enable a quantum leap in granularity looking at the U.S. banking industry.

Coupled with our analyzers, it’s possible to generate analysis on bank units as they file their CALL’s and, based our findings, generate a industry picture in around 36 to 48 hours after the close of the CALL submittal window. That’s pretty good!

Stable Pie Slices, But Dramatic Increase in Banking Industry Stress

Based on looking at sample set of around 91% of the test population used in Q4 2008, we observe that the distribution of IRA Bank Stress Rating grades for Q1 2009 retains the roughly 2/3rd to 1/3rd ratio of banks with A+/A grades versus elevated stress institutions IRA detected was characteristic of the banking industry throughout 2008.

Q1 2009 Preliminary IRA Bank Stress Rating Grade Distribution
(Based on data for 7,519 bank units from the FDIC CDR*)

Source: FDIC/IRA Bank Monitor
IRA Bank Stress Grade2009 Q1* 2008 Q4
A+ 3362 3918
A 1909 1705
B 135 119
C 411 390
D 87 98
F 1615 2003
There were an additional 108 banks in the initial data set that were found to have some holes in the data thus preventing computation for the preliminary analysis run.

We see indicators of a continued migration of banks from the A+ range where stress fall below the index Dec-1995 = 1.0 start mark into the A range indicating more banks are now feeling the effects of economic conditions regardless of the business practice models they’ve had in place.

At this time we do not know what the disposition of the remaining 750 or so institutions that are not in the CDR as of 5/3/2009. We have not yet had a chance to determine who these missing units are.

IRA Historical Bank Stress Grade Distributions
based on FDIC Research Master Files

Source: FDIC/IRA Bank Monitor
Period A+ A B C D F
2008 12 3,918 1,705 119 390 98 2,003
2008 09 4,498 1,325 283 356 63 1,793
2008 06 4,884 1,248 404 326 66 1,458
2008 03 5,167 1,042 578 334 68 1,233
2007 12 5,556 610 884 315 70 1,029
2007 09 5,931 395 950 274 37 902
2007 06 6,056 354 972 273 60 824
2007 03 6,075 304 1,057 284 63 795
2006 12 6,370 134 1,165 204 39 697
2006 09 6,666 29 1,108 198 44 628
2006 06 6,729 0 1,155 194 35 613
2006 03 6,752 2 1,131 187 39 608

At first glance the situation as of Q1 2009 may seem to be getting better. But it’s not. In prior quarters, banks wound up getting “F” grades because they were barely making money; that is, they had small but positive net incomes that produced ROE’s sufficiently below industry averages to indicate elevated business operating stress. A lot of these institutions were suffering due to mark-to-market accounting, goodwill write-downs and other ROE issues.

In Q1 2009, the data indicates a dramatic climb in the industry aggregate average Bank Stress Index from 1.8 at the end of Q4 2008 to a whopping 5.57 coming out of 1Q 2009 or half an order or magnitude above the 1995 benchmark. The reason for this is the number of banks who delivered negative net incomes in the first quarter of 2009, one thousand five hundred fifty-seven (1,557) of them as updated on our system after the data run on May 5, 2009.

Keep in mind what the Q1 2009 FDIC data is saying: Even with the change in the FASB rule for M2M accounting, and Fed liquidity programs, the leading factor driving higher industry stress scores remains ROE degradation, not charge-offs or operational factors such as efficiency. When charge-offs are the leading factor in the IRA Banking Stress Index, then the industry will be through the worst.

Number of Bank Units with Negative
Net Income in 1Q2008 by State

updated as of May 5, 2009
Source: FDIC CDR/IRA Bank Monitor
AL 27
AR 11
AZ 39
CA 135*
CO 27
CT 12
DC 4*
DE 11
FL 149*
GA 139*
HI 1
IA 29
ID 5
IL 108*
IN 10
KA 1
KS 37
KY 20
LA 11
MA 27
MD 16
MI 53
MN 77
MO 64*
MS 7
MT 10
NC 36*
ND 10
NE 27
NH 4
NJ 26
NM 5
NV 23
NY 30*
OH 22
OK 19
OR 14
PA 41
PR 2
RI 3
SC 18
SD 8
TN 33*
TX 89*
UT 23*
VA 20
WA 49*
WI 22
WV 1
WY 2
* items updated since 5-3-2008. Please note that th FDIC CDR system continues to release data. The definitive lock down of quarterly numbers happens when the research masterfile is released later this month.

The Q1 2009 results calculated by IRA are looking a little like a table from the CDC’s H1N1 confirmed laboratory cases page. Our overall observation is that U.S. policy makers may very well have been distracted by focusing on 19 large stress test banks designed to save Wall Street and the world’s central bank bondholders, this while a trend is emerging of a going concern viability crash taking shape under the radar.

We’ve noted in the past that US banks have been migrating down the quality slope taking an average of 9 months to complete the journey from “A” to “F” on the stress scale. The story is predictable. It begins with business losses and recriminations. This is followed by lending engine contraction as the propensity to create exposure narrows towards risk aversion and “quality lending” exclusivity. This is a rare diet to try to live on in these times. The bank, which makes its’ living wage from the interest it collects from its’ lending engine slowly starves.

At a certain point the carry cost of the infrastructure outweighs the earnings rate. Then you start to see strange shifts to seek incremental income from service fees, a move that often only serves to increase customer reluctance and mistrust. The end result is a stressed business model that can only be remedied by getting the core business, its’ lending engine, running again.

Counting the fourth quarter of 2007 when IRA’s data indicates this phenomenon began to emerge, we are now eighteen (18) months or one-half of a business cycle dragging this massive boat anchor the behind the US economy. We may have wasted valuable time trying to save Wall Street at the cost of Main Street.

At this point the reasons no longer matter. It’s time to win by thinking gracefully. The numbers indicate we need to seriously ask the question as to whether economic recovery for the United States can still come just from repairing Wall Street – or whether instead we should be worried about addressing the underlying loss rates that are driving the provisioning behind these poor ROE results. Has the time come to shift the policy focus away from the things that we love, namely big zombie banks, to tackle things that are truly hurting us?

Readouts on all 7,519 bank units collected by IRA to date are now available to our Advisory Clients. The Beta version of unit level preliminary indicators now appears in the IRA Bank Monitor for Professionals. Preliminary grade indicators appear in pink if available. It will begin to appear in the IRA Bank Ratings Service for Consumers as soon as Beta testing is completed. IRA is presently working on adapting the accompanying bank-holding company (BHC) extension of our ratings system to also feed from CDR collected preliminary data. We continue to support our “prime solution” philosophy that it takes everyone with fair, equal and transparent access to risk information to collectively guide our economy to recovery.

1 comment:

  1. Am I nit picking if I sense a disturbance in the force between this post and the previous? In one place, you applaud the FDIC for providing a timely public data repository, which I understand it can only provide thanks to its streamlined XBRL processes, and in the other place, you applaud the new SEC Chair for her scepticism towards a speedy implementation of the same? What gives?