Remember the issue of declining coverage ratios at European and US banks?
Coverage ratios are essentially loan loss reserves (provisions for bad debt) divided by non-performing loans, and they declined in the second-quarter for European banks and the third-quarter for US banks, collectively — which basically means banks were provisioning less for their bad debt.
That could be fine if there isn’t a significant increase in non-performing loans (NPLs) in the future — but it could prove to be a concern if there is.
And if a piece of research from Jason Goldberg, at Barclays Capital, on US banks’ third-quarter results is anything to go by — it seems regulators are on the “concerned” side of the equation. Here’s Goldberg:
In 3Q09, reserve/NPLs fell from 85% to 82%. While not our favorite metric (remember that in several circumstances NPLs are already written down to market value) and it is not uncommon to see this metric fall at this point in the cycle, we believe that some banks have to be approaching ‘floors’ and believe regulators are looking at this. Only 4 banks (CMA, MI, USB and KEY) saw this metric improve linked quarter, with MI, STI, HBAN, RF, SNV and ZION all under 60%. Credit card heavy (low NPA category), JPM, BAC and C, as well as EWBC, WFC, KEY and NTRS were over 100%. While REO expenses jumped, assets declined 2% at the median bank (composite up 4% however). CYN, NTRS, USB, BBT, ZION and RF saw increases in excess of 15%, while HBAN, BK, SNV, EWBC, C and MTB witnessed the largest declines. Restructured loans continued to increase, while OREO trends were more mixed.
In graphic terms — that decline in Q3 coverage ratios looks like this:
Now if European banks’ second-quarter results have taught us anything, it’s that fluctuating coverage ratios have an impact on earnings. To put it simply, if the coverage ratio is lower, earnings tend to get a boost — albeit at the potential expense of future earnings if loan losses pick up again. If the ratio’s higher, then earnings tend to take a hit — but the bank is (presumably) better provisioned to deal with future loan losses.
So if regulators start looking at the measure or enforcing some sort of minimum ratio, it could mean an additional drag on bank earnings.
As a bonus, and while we’re on the subject of non-performing loans, provisions, etc., here’s a bit more detail on Q3 US bank results from the BarCap note.
To start with — non-performing assets:
NONPERFORMING ASSETS — While the rate of growth did slow, every bank except for two (HBAN, MI) saw NPAs increase linked quarter. For our composite, NPAs increased 15% or $21 billion to $162 billion (vs. our forecast of $160B). The rate of growth slowed for the 3rd straight quarter (vs. 21% in 2Q, 29% in 1Q, 36% in 4Q). The median bank saw its dollar NPAs increase 14% linked quarter, down from 19% in 2Q09 and 29% in 1Q09. For our coverage, 70% of the names saw their percentage rate of increase in dollar NPAs slow in 3Q09 relative to 2Q09 (C, BBT, BK, EWBC, HBAN, TCBI, ZION the exceptions), with BAC, FITB, KEY, PNC, USB, NTRS, CYN, MI and MTB witnessing the pace slow in each of the past two quarters. MI and FHN saw modest declines, while SNV, KEY, STI, CMA, TCB and BAC put up single-digit increases. BK, TCBI, ZION, NTRS, WFC, BBT, EWBC, RF and PNC all witnessed increases of 20% to 50%. Every bank had positive NPA formation ((EOP NPAs –BOP NPAs + NCOs) / BOP NPAs) this quarter, led by EWBC, BK, JPM, WFC and TCBI, all over 50%. MI, FHN, STI, MTB and TCB were all under 25%. Relative to the prior quarter, MI, RF, KEY, STI and TCB posted the sharpest slowdowns, while BK, EWBC, TCBI, SNV, C and HBAN witnessed an acceleration.
In 3Q09, the median bank’s NPA ratio increased 52bps, only modestly better than the very large increases of 54bps in 2Q09 and 65bps in 1Q09. While many did call for the pace of the increases to continue to slow, we had hoped for more of an improvement, though note that a greater than expected drop in loan balances did influence this metric. Still, we are running at extremely elevated levels. While still compiling all the granular data, there did seem to be a shift in the drivers on NPA growth, from the consumer side to commercial real estate. Additionally, for those that disclosed it, increased restructured loans also played a role.
Still, the median bank’s NPA ratio was 3.70%, slightly better than our 3.80% projection. Still, this compares to its 10-year average of 0.60% and ahead of the 2.70% level seen in 1991. Every name under coverage experienced a linked quarter increase in its NPA ratio with ZION (up 198bps), RF (112), FITB (91), HBAN (90), TCBI (86) and C (84), all experiencing increases of over 80bps. STT (nil), TCB (up 12bps), MI (19), NTRS (27), USB (27), MTB (28) and FHN (29) stood out at the other end seeing increases of less than 30bps. HBAN (7.2%), MI (6.8), ZION (6.6), SNV (6.6), FHN (6.6) and STI (6.4) all had NPA ratios over an alarming 6%, in part due to a continued concentration of construction and development credits. It is unclear to us how banks will be able to work down all this supply near-term. The trust banks followed by TCB (2.5%), EWBC (2.7), TCBI (2.8), MTB (2.9) and WFC (2.9) are the only names under 3%. Recall, two years ago every name under coverage was under 1.25%.
Still, we remain concerned NPAs may be understated for some, particularly because several banks exclude and/or don’t report restructured loans, as well as TDRs, which we expect to have high re-default rates. Furthermore, others are extending the maturity dates of certain residential real estate related loans, namely construction and lot loans, as well as using interest reserves. With respect to construction loans, we continue to highlight those names with exposure that exceeds 275% of tangible equity, including UCBH, SNV, ZION, BBT and MI, a level at which we remain concerned.
Of the names that report it, TCBI, SNV, FITB and STI saw the largest increases in loans 90-days past due, while KEY, HBAN, CMA, PNC, MTB and MI saw double-digit declines. Still, we have concerns that the data is being skewed. We highlight Citi’s 2Q 10-Q filing stated that while it observed declines in its 90 to 179 day mortgage delinquency bucket, well over half of the decline was attributable to loss mitigation and modification initiatives.
And on net charge-offs (loans written off as uncollectable):
NET CHARGE-OFFS – For our composite, NCOs increased 7%, or $2.4 billion, to $38.7 billion (vs. our estimate of $40B), a sharp slowdown from the 27% increase last quarter. The median bank’s NCO ratio increased 30bps linked quarter (vs. our estimate of up 25bps), down from last quarter’s 63bps rise. Similar to NPAs, the pace of net chargeoffs increases slowed, though we are very mindful that the increase in 2Q09 was the second largest increase in history. The median bank’s NCO ratio was 2.88 %, which is more than triple 1991′s annual level of 0.88% and compares to a 10-year average of 0.38%. This is up from 2.10% last quarter, and above our estimate of more than 2.60%. Since the start of 2008, our banks have charge-off $175 billion of loans. This represents roughly 5% of the average loans over this seven quarter span. Seven banks (FHN, TCBI, MI, PNC, C, BBT, MTB) witnessed improvement in their NCO ratios, up from just one last quarter. SNV (up 223bps), RF (82) and STI (78)–all Southeast based–followed by FITB (70) and KEY (70)–both Midwest–posted the largest increases. SNV (7.3%) and EWBC (7.1%) both posted NCO ratios over 7%, while C (5.0), JPM (4.9) and BAC (4.2) due to card and MI (4.5) and FHN (4.2) due to construction were over 4%, while the trust banks and TCBI (0.19%), MTB (1.1%) and TCB (1.5%) continued to stand out on the other end.
While still aggregating credit quality data by type, our initial thoughts are home equity did better than expected, credit card was in-line and commercial real estate, and to a lesser extent C&I, did worse. Residential construction related write-downs continued to be very burdensome and we look for this early part of the cycle to have a tail. Additionally, deterioration is spreading into commercial construction, which is a bigger asset class. CRE got a lot more air time this earnings season, as problems spread beyond homebuilders, and deeper into the likes of multi-family, retail, and lodging and even office. C&I also showed increased signs of stress, beyond businesses tied to housing and toward economic-sensitive industries (retail, media), amid an increase in bankruptcy filings, though we believe this asset class remains manageable.
Lower property values, increased vacancy rates, and increasing refinancing risks continue to drive our CRE concerns. Property valuations have dropped 35-45% from peak levels in 4Q07, driven by the sharp decline in transactional activity. Also, lower occupancy rates and rents are increasing the number of properties that aren’t generating enough cash to make principal and interest payments. Lower values could make it more difficult for borrowers to be able to extend existing mortgages or replace them with new debt. We believe banks have been extending CRE debt when it has matured. Still, with more CRE loans coming up for refinancing, the system’s capacity appears limited amid increased vacancy rates for retail and office properties and a decline in asset prices, which could precipitate non-residential foreclosures (though watching TALF and PIPP with a hopeful eye). We believe there is over $3.5 trillion of outstanding debt associated with CRE, with almost half ($1.6B) held on the books of banks and an additional $900 billion representing collateral for CMBS. In our coverage, EWBC, SNV, ZION, UCBH, BBT, MI, CMA, and RF all have over one-third of their loan portfolios in CRE/construction loans. In addition to asset quality, we also note real estate has been a significant source of earnings growth for this group, and while housing appears to be showing signs of stabilization, we are not looking for a snap-back, which could impact future revenue streams. Fortunately from a systemic standpoint, real estate concentrations tend to be focused in the smaller, regional players. The bigger banks’ loan losses are more skewed toward the consumer, while the regional banks’ losses tend to be more commercial real estate heavy.
Related links:
Great Depression-esque bad debt at US banks – FT Alphaville
Banks’ coverage ratio capers, cont. – FT Alphaville
Good banks, bad banks – FT Alphaville