In general, short interest in large US banks and thrifts is small, and has been so for quite some time. Short interest in JPMorgan Chase (JPM), the largest US bank by assets, was only 1.2% of shares outstanding as of May 15th, the most recent data available, and since mid-September 2007, the earliest date for which SNL Financial provides such data, has never been higher than 1.9%. Other institutions have seen their short interest spike, in some cases to 20% or more, and then fall. For example, Citigroup's (C) got as high as 23%, before declining to its current 1.3%.
Currently, of the 21 institutions with assets over $30 billion, five institutions have short interest over 7%, versus a group median of 1.5%. The five institutions are listed below, along with relevant background information:
M&T Bank Corporation
New York Community Bancorp, Inc.
People's United Financial, Inc.
Why is short interest in these names so high? Do the short positions appear to be driven by valuation concerns (high valuation relative to operating performance), asset quality concerns, or both? And do the shorts have a point? I'll discuss each name individually below.
M&T Bank Corporation (MTB) - Of the five institutions listed above, Buffalo, NY-based bank MTB has the strongest operating performance. Since the beginning of 2000, it has grown tangible book value per share ("TBV-PS") at the fastest rate, 8.7% compounded annually, and its 2012 EPS of $7.54 is the highest the bank has ever delivered. Sell-side analysts expect meaningful growth in 2013 and 2014. Through the financial crisis, MTB never posted a quarterly loss. While MTB has occasionally made acquisitions, its growth has largely been organic. 2012 EPS for three of the other four institutions are well below their respective highs, and the fourth institution has grown EPS at a painfully high cost to shareholders. MTB's historical return on average tangible common equity ("RoATCE") has consistently been the highest, in the 30% range. While RoATCE has more recently been lower, around 20%, this appears to be due mainly to a material recent increase in tangible common equity/tangible assets ("TCE/TA"), from an historical level of about 5.5% to 7.48% at Q1 2013.
MTB's valuation reflects its superior operating performance. Of the five institutions, MTB and NYCB have traded at a consistently higher price/TBV-PS multiple than the other three. MTB closed on Friday at $103.74, or 2.26x TBV-PS, which doesn't seem unreasonable given its RoATCE and solid historical growth.
Is asset quality a concern? Like many banks, MTB saw its non-performing assets ("NPAs")/total assets spike relative to historical levels. They peaked at 2.31% in Q1 2010. They've fallen materially since then, to 1.64% at Q1 2013, but they haven't fallen as rapidly as they have at ZION, or to as low a level as they have at CMA or NYCB. And they're still well above historical norms.
On August 27th, MTB announced that it would acquire Paramus, NJ-based thrift Hudson City Bancorp for $3.8 billion. The deal hasn't yet closed, and therefore the above numbers do not reflect Hudson City's results. MTB is paying 0.85x TBV-PS. While that may seem like a steal, Hudson City's RoATCE is very low, in the 4-6% range. Furthermore, its NPAs/assets at Q1 2013 were 2.93%, materially higher than MTB's 1.64%. And in contrast to MTB's declining NPA/asset ratio, Hudson City's has been rapidly increasing. Hudson City has $40 billion in assets, so it is roughly half MTB's size. It seems likely that short sellers are interested in MTB because of this deal.
MTB's short interest got as high as 16.5% in mid-March 2009, and spiked again to 15.1% in October 2010, before falling to a low of 2.3% in May 2011. M&T's price rose steadily after March 2009, so it appears that the shorts took a beating. Short interest began increasing again at the end of July 2012, from 3.5%. Since that time, MTB's price has increased by 21%.
So some investors seem to be big fans of the Hudson City deal, and others are skeptical. Who is right? Given MTB's solid historical operating performance, strong management team and disciplined approach to acquisitions, I'd be hesitant to bet against it.
Comerica Incorporated (CMA) - The operating performance of Dallas, TX-based bank CMA has been a mixed bag. CMA's EPS in 2001 through 2004 were all below 2000 EPS. However, from the low of 2002, EPS grew steadily. CMA hadn't been an active acquiror for many years, although that changed in January 2011 when it announced it would acquire Houston, Texas-based bank Sterling Bancshares for $1.0 billion. Prior to the financial crisis, CMA's RoATCE had been in the 15-17% range. Depressed earnings and reduced TCE/TA motivated CMA to issue common equity in March 2010, which diluted lower post-recovery earnings. CMA earned $2.67 in 2012. While this is a solid improvement over the $2.09 of 2011 and the $0.79 of 2010, it is still well below the 2006 peak of $5.50. RoATCE in recent quarters has been just above 8%.
CMA closed Friday at $38.86, or 1.15x TBV-PS. This is a low multiple, but is arguably a bit high given CMA's low RoATCE. CMA had traded above 3x TBV-PS in 2000 and 2001, but it hasn't traded above 1.5x TBV-PS since October 2007.
CMA's NPAs/assets peaked at 2.47% in Q3 2010 and have since fallen to 0.99% in Q1 2013. Net charge-offs ("NCOs")/average loans hit a near-term peak of 2.13% in Q3 2009, and have since fallen to 0.21%. Reserves/NPAs appear solid. The Sterling deal does not appear to have worsened CMA's asset quality.
CMA's short interest got as high as 17.3% in July 2008. It fell considerably from there (share price did too - it looks like shorts made money here), to 4.69% in April 2009, and then gradually crept up to 9%. It increased from a near-term low of 4.1% in mid-December 2012 to its current level. CMA's asset quality was steadily improving over this period, but CMA's share price increased by 36%. This implies that shorts are focused on CMA's valuation rather than its asset quality.
CMA's stock price doesn't appear likely to collapse, but it could fall 10% or more from its current level.
New York Community Bancorp (NYCB) - I wrote a detailed report about Westbury, NY-based thrift NYCB for Seeking Alpha in March. NYCB has been highly acquisitive, and that has helped it grow assets rapidly, from $1.99 billion at Q1 2000 to its current $45 billion. However, growth in assets per share ('A-PS"), the better asset growth measure for a serial acquiror, was 42% on an annually compounded basis from Q1 2001 to Q1 2004, but has been only 0.4% since. TBV-PS grew more rapidly than this, but that was because of an $897 million common equity issuance in December 2009. NYCB's peak EPS of $1.65 occurred in 2003; it earned $1.14 in 2012. In 2002 and 2003, RoATCE was often a nosebleed 40% or higher, helped by a 2.00%+ return on average assets ("RoAA") as well as a TCE/TA of 4.00% or lower. More recently, NYCB's RoAA has been around 1.10% and its TCE/TA around 7.6%, making current RoATCE a much lower, although still high, 15-17%.
At points in 2001 and 2003, NYCB's price/TBV-PS multiple was over 7x, presumably because of its exceptionally high RoATCE combined with a solid growth trajectory. This multiple declined gradually over time. The last time it was consistently over 2x was in mid-2011. On Friday, NYCB closed at $13.22, giving it a 1.82x TBV-PS multiple. A high multiple in absolute terms, but not high given NYCB's 15-17% RoATCE.
From an asset quality standpoint, NYCB looks a bit better than MTB and CMA. So why is short interest so high? Short interest spiked to 14.0% in March 2009, before declining over time to 3.2% in mid-August 2011. Short interest has risen since then, although share price has moved sideways. For the last several years, NYCB has paid out most of its earnings as dividends. Its 2012 dividend payout ratio was 88%. In 2008, it was 455% (cutting this to "only" 100% would have saved $270 million, 30% of what NYCB had to raise a year later). The stock sports a 7.6% dividend yield, way above the more typical 2-3% bank/thrift dividend yield. NYCB is adequately capitalized, so there's no reason to expect NYCB will cut its dividend. But note that NYCB's median sell-side EPS estimates for 2013 and 2014 are $1.02 and $1.00, respectively. Yes, the 2014 estimate is lower than 2013's, and both are meaningfully below 2012's actual EPS of $1.13. That means the dividend payout ratio will ratchet up to 100%.
NYCB CEO Joseph Ficalora has expressed an interest in pursuing additional acquisitions, not a positive given the thrift's acquisition track record. If a thrift pays out 100% of its earnings as dividends, it's admitting it can't grow organically. Why should such an institution chase deals?
The shorts seem to have good reasons for liking NYCB.
Zions Bancorporation (ZION) - Salt Lake City, UT-based bank ZION delivered solid asset growth (most of it organic) and EPS growth from 2000 through 2006. Its asset quality began deteriorating sharply in mid-2007. NPAs/assets rose steadily from 0.20% to a very high 5.2% in Q1 2010. None of the other four institutions profiled in this report saw peak NPAs/assets of even 3%. ZION lost money in 2008, 2009 and 2010. Because of an especially large loss in 2009, it issued common equity and ultimately increased shares outstanding by 60%. ZION is still materially less profitable than it had been; 2012 net income was $348 million (admittedly depressed by a $74 million pre-tax realized loss on securities), versus a peak of $583 million in 2006. ZION's higher share count makes the EPS comparison even worse. ZION's 2012 EPS was $0.97, versus $5.37 in 2006. Median EPS estimates for 2013 and 2014 are $1.80 and $1.90, respectively. ZION can hit these numbers if loan loss reserve releases continue and no large unanticipated charges arise.
ZION's asset quality is clearly improving. NPAs/assets have fallen from their 5.2% peak to 2.0%, about even with PBCT but still higher than MTB, and materially higher than CMA or NYCB. On Friday, ZION closed at $27.37, at a price/TBV-PS multiple of 1.26x.
ZION's short interest peaked at a nosebleed 41% in mid-September 2008. It fell sharply from there, to 18% in December 2008, just as ZION's share price was in the early stages of what became a 70% price decline. The next major short interest decline began in mid-November 2009, from 25%. By mid-May 2010, short interest had declined to 11.7%, and it has gradually declined to its current level since then, with the stock price holding steady over this time period. ZION's stock price has increased by 23% since mid-October 2012, but short interest hasn't declined appreciably.
ZION's 2012 dividend payout ratio was only 4%. If ZION does hit the median EPS estimates in 2013 and 2014, its TBV-PS will grow from its current $21.67 to $24.33. ZION's 2014 implied RoATCE is about 8%. ZION seems pricey, unless you expect major improvement in its RoATCE. The sell-side isn't expecting any such improvement.
People's United Financial, Inc. (PBCT) - I wrote a detailed report about Bridgeport, CT-based thrift PBCT for Seeking Alpha in April. To summarize briefly, PBCT's operating results have been dreadful, mainly because of poorly structured and/or executed acquisitions. PBCT's TBV-PS hit its high of $14.72 in Q3 2007, thanks to its second-stage conversion from a mutual holding company to a fully-public institution. TBV-PS had been $4.20 only two quarters prior. Because it used cash as consideration in acquisitions and repurchased shares at inflated levels, all in an effort to solve the "problem" of excess capital, PBCT ran TBV-PS down to $8.33 at Q1 2013. Through the financial crisis, PBCT never incurred a quarterly loss, so this decline wasn't caused by losses or by the necessity of a dilutive common equity issuance.
PBCT closed on Friday at $13.92, or 1.67x price/TBV-PS multiple. That's a nutty price for an institution with an 8-9% RoATCE. The reason it trades so high? The same reason NYCB does. PBCT sports an above average 4.7% dividend yield. In 2012, it paid out 89% of its EPS as dividends. In 2008, 2009 and 2010, it paid out well over 100%. PBCT is still somewhat over-capitalized, with a TCE/TA ratio of 9.6%, so there's relatively little risk that PBCT will cut its dividend in the near term.
The median sell-side estimates for PBCT for 2013 and 2014 are $0.78 and $0.88, respectively, versus $0.72 in 2012. So the Street expects EPS growth of 8% and 13%, respectively. Where will it come from? Buybacks? Another acquisition?
PBCT's NPAs/assets peaked at 2.8% in Q4 2010, a bit later than the other four institutions. At Q1 2013, NPAs/assets stood at 1.93%, a bit high and basically on par with ZION. Reserves/NPAs are currently 31.8%, the lowest of the five. The next highest is NYCB with 58.8%.
From the standpoint of short interest, PBCT has been relatively boring. Since September 2007, PBCT's stock price has trended down slightly. Short interest spiked to 7.6% in mid-February 2011. PBCT had announced its $489 million acquisition of Danvers Bancorp on January 20th. Short interest then fell rapidly to 2.3% in mid-August 2011, with no meaningful share price move. The recent upward trend from 4.1% in mid-November 2012 to the current 7.1% was accompanied by a 20% share price increase.
PBCT has expressed an interest in pursuing more acquisitions. That, the recent share price run-up and its current price/TBV-PS multiple are three good reasons to short the stock.
As large as the short positions are in the five institutions I profiled, some smaller ones have short positions of 10% or higher. I'll publish a companion piece to this one discussing them sometime soon. And if there are other topics in the bank space you think deserve attention, don't hesitate to send them along.