Greetings StockCentral Community!
Today I hope to go beyond the panic of newspaper headlines about the real estate market and try to put some real numbers around the EWBC loan portfolio. Let’s see what stress tests we can perform and exactly how bad a real estate recession could be for EWBC.
As pointed out earlier, real estate prices seem to be one of the key risk factors in holding EWBC.
I used the following resources:
1. There’s a report available from the FDIC, the Uniform Bank Performance Report (UBPR), that gets us down into some terrific detail about any bank’s loan portfolio. This report is available at http://www4.fdic.gov/ubpr/ and can be viewed on screen or exported into a format the Microsoft Excel can import. Unfortunately, data only goes through yearend 2007 as of today.
2. EWBC’s Q1 2008 earnings report and financials at http://biz.yahoo.com/bw/080415/20080415006744.html?.v=1 This includes a table of Q1 loan delinquencies.
3. The EWBC Q1 earnings call transcript at http://seekingalpha.com/article/72597-east-west-bancorp-q1-2008-earnings-call-transcript?source=yahoo
I have attached an Excel spreadsheet which contains loan amount and delinquency data (and my analysis) derived from the UBPR and the earnings report. Unfortunately, the data from the two sources were categorized somewhat differently, so I could not consolidate the two.
In the spreadsheet I itemized the various types of loans on the EWBC balance sheet and indicate the dollar amounts outstanding and delinquency rates outstanding on each loan type.
One of the tabs looks at delinquency rates of yearend 2007 versus yearend 2005. I did this to compare delinquency in today’s environment versus a “normal” economy, in order to identify the loan categories that deserve special analysis. If the delinquency rate does not change very much, I’m assuming that there’s not a problem and that the delinquency rate will stay around the same going forward. For example, this seems to be the case for commercial and industrial loans.
The other tab compares delinquencies in Q1 2008 versus Q4 2007, to see the latest trend.
By the way, I am defining “delinquent” as any loans whose payments are more than 30 days late. This is a pretty aggressive definition of delinquent. I use this broad definition on the belief that it provides the best “first glimpse” of a delinquency issue, even if an actual delinquency may be unlikely. Historically, actual chargeoffs are only a small fraction of delinquencies, as you can see in the UBPR.
Note that there are some categories where there is no delinquency data for 2005. There are also some loan categories that were so small I didn’t worry about them.
From the UPBR report, the three problem categories seemed to be
- Construction and Land Development: delinquency went from 1.84% to 2.16% based on $2B of loans.
- 1-4 Family Residential: delinquency went from 0.3% to 2.54%
- Loans to Individuals: delinquency went from 0% to 1.73%. The total loan amount was small, though.
By way of “stress testing,” let’s assume that these three above-mentioned delinquent balances are worth nothing (i.e. no resumption of payments and no recovery). The total writeoff would be $57M, or 22% of 2007 pretax profit.
If one doubles this amount, the total writeoff would be $115M, or 44% of pretax profit.
In either case, these writeoffs, if they are a one-time event, would be covered by current earnings. By way of comparison, 2007 pretax profit was $262M.
On another tab, I performed a similar analysis using the Q1 2008 data. If I doubled the delinquent loans and assumed there were fully written off, the resulting figure of $225M would just about wipe out a year of pretax profit. Not a great thing, but still, if limited to a one-time event after which EWBC resumes normalcy and growth, the stock could perform well in a 5-year timeframe.
In the earnings call, management discussed their conservative loan-to-value ratios. The average for residential construction was 69% on average, commercial construction was 55% on average, and land was 45% on average. Assuming that the original appraisals were not over-aggressive and that these averages don’t hide a wide distribution (e.g. if 20% of the loans were at a 90% LTV, this could be a problem), then this would seem to offer protection against total writeoffs in the event of delinquency. It’s on my list to ask investor relations more about this. Management also pointed out that low LTV ratios make it less likely the property owners will walk away.
There was some analyst discussion about loans to the Inland Empire region of southern California. I live in Boston, and this reminds me a little of outlying areas of Boston, like Framingham, during the last real estate bust in 1989-1991. The area was hardest hit during the bust (one could buy apartment buildings there for $10K - $20K per unit), but got one of the greatest percentage increases during the boom. I suspect Inland Empire is similar, with great opportunity during a boom but first and farthest to fall during a bust.
According to the earnings call, only 11% or $116M of EWBC residential construction loans were in Inland Empire. $145M in land loans were in Inland Empire. It does not seem like huge exposure given the total loan portfolio of nearly $9B, and assuming that only a portion of these loans will go south, and that of that, some amount will be recoverable.
That said, I would like to ask IR the amounts of other loan categories tied to Inland Empire real estate.
Overall, the data is giving me some comfort on the risk level. What am I missing here?
|