Industry
After the collapse of Silicon Valley Bank in the US (and then Credit Suisse in Europe), regional banks (KRE) have now declined by 44% from their January 2022 highs, in line with past recessions.
According to research from Bank of America, in seven recessionary scenarios (excluding the Great Recession), banks (the BKX Index) have rallied on average 34% over the six months following a 40% decline.
In the following year, bank stocks have rallied on average 56% after a 40% decline in the BKX.
If this isn’t the Great Recession, then our odds appear quite high that financials will recover. Compared to the situation back then, banks today have more capital, better underwriting standards, and far better liquidity.
In the Great Recession scenario, by the way, the BKX fell 80% from peak to trough, so there is a risk should there be a similar credit risk meltdown. But again, the structural leverage and credit problems at banks then were well beyond what they are today.
M&T Bank
M&T Bank (NYSE:MTB) is a top 15 bank in the US with a $20 billion market cap. It was founded in 1856 and is one of the most conservatively run banks in the country. As per this slide, their branch network has been built for decades on local relationships (very different from private equity/California-based start-ups who have fast money depositors).
M&T has had zero money-losing years dating back to as far as I could find data, in this case, 1987. In 2009, EPS fell 40% from $5.42 to $3.21 per share. At the time, their Tier 1 common equity ratio was 5.7%, a number in today’s world that would be considered unacceptably low (and risky). Today, the company sports a solid balance sheet with Tier 1 common equity at 10.44% and only a small amount of Held To Maturity assets (6% of the balance sheet).
They are one of only two banks in the S&P 500 that never cut their dividend during either the Great Recession or the pandemic. Berkshire Hathaway owned 5.3 million shares of M&T until the pandemic struck in 2020, at which point they sold out by year-end (at prices probably 10-20% lower than today even).
Longer term, M&T has grown EPS at a 6.0% CAGR since 2007 (a peak year) through 2022. The dividend yield is approaching highs not seen except during the pandemic and the Great Recession.
Below is the stock's dividend yield since 2005.
Since 1997, M&T has only raised its dividend, barring a few years where it was kept flat. They purchased People United (PBCT) in April 2022, which was a Dividend Aristocrat.
Below is a slide pointing out M&T’s returns on equity (solid at 12%), credit metrics (with little in the way of charge-offs), conservative loan-to-deposit ratio (80%), and efficiency ratio (53%). Tier 1 CET capital was 10.44% at year and estimated to be 10.15% at the end of March. We get why Buffett has been a shareholder in the past.
Above you can also get a glimpse into the company’s allowances for loan losses vs net charge-offs. That is, the company has loan loss reserves of 1.46% of total loans, but only charged off 0.13% of loans in 2022. Allowances incorporate future loan loss expectations, of course, but this bank is well known to be one of the best underwriters in the U.S.
In all economic downturns above, M&T has shown best-in-class net charge-off (NCO) results.
Today’s Environment
Right now, investors are hyper-focused on 3 things:
- FDIC guarantee levels, deposits and the liquidity available to fund a bank run,
- Duration mismatches, that is, banks that hold high levels of Held To Maturity assets as a percentage of book value, and
- Commercial real estate exposures, which are particularly worrisome on the West Coast and especially in the New York/Northeast region.
Starting with deposits, we post this from the 2022 10-K.
While the 10-K does not disclose the level of insured deposits, we can guess that most of the retail deposits are small (with M&T operating over 1,100 branches), with commercial accounts likely over the limit. That suggests that about half of their deposits are uninsured.
Banks in the upper right quadrant of course are most susceptible to a bank run / need for capital. Those like MTB in the lower left quadrant are the least likely to see any kind of deposit flight. This chart also suggests roughly 50% of deposits are insured.
MTB’s long history and relationship-oriented banking approach should keep deposit outflows to a minimum. They have a lot of branches, scale, and a perception of conservatism in their core territory. Given that trust is critical today, we view MTB as high on the trust scale.
Liquidity levels are quite high, too. Against $159 billion of deposits as of March 31st (down only 2.7% from year end), MTB has cash and bank deposits of their own of $24 billion. They also have $10 billion of easily saleable Held For Sale (HFS) securities, not to mention Fed financing now available on their HTM assets.
The company is seeing cash migrating from non-interest-bearing checking accounts to interest-bearing accounts (as all banks are). They also cited competition from external sources of higher rates (with Treasury Direct mentioned). There were also outflows of $700 million in deposits in the first quarter as escrow deposits fell (a typical seasonal phenomenon).
For the full year, M&T expects total deposits to fall in the low single digits.
With other lending facilities from the Federal Reserve Bank of NY and the FHLB, M&T has another $34 billion of borrowing capacity.
This adds up to total liquidity of $85 billion (versus deposits of $159 billion). The company would have to see deposit outflows of over 50% of the total before they would run into problems (no pun intended). Again, last quarter they were 2.7%.
Below is a look at M&T’s investment securities (as of year end as the March 2023 10Q has not been filed yet).
Compared to SIVB, which owned HTM assets at a 1.7% yield, M&T waited to deploy cash until well into 2022, putting money to work at far more attractive yields (highlighted above at 2.92%). With $23 billion of their own deposits sitting at other banks, they intend to deploy more capital ($4 billion per the last call) into longer-duration assets and pick up some net interest income to offset NIM percentage pressure.
Overall, our view is that M&T’s interest rate/duration risk is low. While confidence in all banks has been shaken, M&T is where the conservative money should be flocking to.
Below is the most recent balance sheet.
As of March 31st, Loans totaled $131 billion against $159 billion of deposits (82%). HTM assets are only 6% of total assets.
Also, Q1 tangible book value per share was $87.40 (slightly higher than year end). Subtracting out the $1.3 billion in HTM unrecognized losses and TBV/share is still a healthy $81.60. MTB historically trades at 2.1x tangible book, which implies $163 on the stock even after taking these losses.
Don’t forget, the bonds will pull to par meaning that the company will regain this lost $5.80 in book value losses over time as well.
Credit Risk
As noted, the history of MTB is one of being a top tier underwriter. While a recession would naturally push loan loss provisions and charge offs higher, we don’t worry too much about anything that might be devastating to the balance sheet or the preferreds.
The trends in loan quality are actually improving slightly over the past year, but nonaccruals did tick up slightly from year end to March 31st (as a % of total loans) by 7 basis points.
We prefer to see Allowances exceed Nonperforming assets (NPLs). Here NPLs are $2.6 billion and allowances $2.0 billion. The company’s response to this is that their recoveries are usually quite high, so even if the full $2.6 billion in nonaccrual loans defaulted, recoveries may end up at 50-75% of par.
That said, a recession appears likely, and so we would not be surprised to see provisions of $500 million to $1 billion in 2023 or 2024.
Our primary concern is with their commercial real estate portfolio, which MTB is actively reducing (since the acquisition of People’s United last year in April 2022). The company has a lot of 2018-2019 construction loans that are getting refinanced/completed and/or moving to mortgages (or to other lenders), so these will decline as the company focuses more on C&I loans (commercial and industrial).
Below is disclosure from the 10-K.
We highlighted above where bigger risks could lie. The company spoke at a conference in early March about their New York City office loans in some detail. Luckily exposure here is low. The overall Office portfolio is $5.2 billion, or 4% of their total loan portfolio (and 12% of the CRE book). NYC office is miniscule and total NYC exposure is 4% of the loan book ($5.8 billion on $129.5 billion of loans).
Commercial construction loans could be risky too, especially in a recession. But we know that tenants prefer newer buildings. We highlight them above as well ($6.4 billion).
On the plus side, MTB noted that retail and hotels are improving today (but created quite a bit of fear in 2020, the last time the stock tanked). Healthcare should also be stable, with industrial/warehouse in good shape.
Investor-occupied commercial loans likely are well underwritten.
While impossible to predict, we can model some economic downturn scenarios. If a modified 2009 case unfolds with bank loans, then provisions could double from current levels. That is a $2.32 hit to EPS (taking provisions up by $519 million).
In a 2020 scenario, provisions might increase by $800 million to $1 billion. That is a $4 EPS hit at the midpoint.
A different approach might be to attempt to mark to market hit to the 2 items above. At 80c on the dollar (where assets in CMBX index land are trading), the impact to book value would be $14 per share.
Below is the latest data on CMBX single A tranche paper.
Note that the quality of MTB’s CRE is probably higher than the index above.
Since the stock has dropped by roughly $65 per share since last October, that seems to more than price in a $14 hit to book from their office portfolio.
As for pricing overall, Capital Economics said that U.S. commercial real estate (CRE) prices have fallen by 4-5% from their peak in mid-2022. They expect a further 18-20% decline. Small and mid-tier banks provide about 70% of outstanding loans to CRE investors.
This from the January earnings call:
In aggregate, that portfolio [CRE portfolio] has still been above 1 [debt coverage ratio]. And when we look at the LTVs in that portfolio, they still run below 60% on a weighted average basis. Now obviously, there are some above that and some below. But as we look at it right now and we look at the clients' ability to support the asset either with cash flow or with how much equity they have in the property where we feel pretty comfortable with where we sit, but we're watching it. As we've talked about, there's -- it's one of the places where we see the most risk and where we're focusing a lot of our attention from a credit perspective."
If CRE prices overall fall 25% from the peak, then given that LTVs were 60% at initiation should still imply 15% loan coverage.
Preferreds
We have been buying these of late, ticker MTB/PRH.
Below is the yield on the preferreds.
We generally would not buy these HIGHER than current prices today ($23.24) and have been accumulating them around $22.
These will float at 3-month Treasuries plus 4.02% in December 2026. If the short end of the curve is anywhere above 2.62%, then the coupon will reset higher and these will be likely taken out. At $22 the YTC is 9.6% a level we deem probably a bottom.
These are only appropriate for small funds or personal accounts given the lack of liquidity in the issue.
What About The Stock?
The company tries to run the business on an interest-rate-neutral basis. The deposit betas compared well to their floating rate book (or asset betas). The company is pretty sophisticated in its approach to duration.
In Q4, EPS was $4.57 compared to $3.50 in Q4 2021. First quarter EPS came in at $4.01, down 7% sequentially but up 53% year over year. With accrued bonus costs hitting in Q1, we deem the “runrate” EPS figure at about $4.25 X 4, or ~$17 per share.
Street estimates today are $16.79 in EPS in 2023 and $16.50 next year. We would guess that provisions in 2024 might knock a couple dollars off of that figure, but then resuming growth in 2025.
Here was guidance as of March 7th.
This year, updated guidance figures suggest 20-23% growth in Net Interest Income (from 23-26% above, driven by declining yields on Treasuries).
If the company takes loss provisions up for their CRE book (let’s say $2.50 including some regulatory fees), and sees some deposit outflows and loan growth of say 2%, then $14-15 is a possible EPS outcome for 2023. That is about flat to 2022 ($14.42 in EPS).
The long term numbers are impressive.
The forward growth numbers are the biggest question mark. This has been a slow growth, mature, boring bank with little material loan growth for years. The stock is more like a bond as we anticipate only 1-4% loan growth in 2023 and 2024. That should improve once past a recession, but this is a re-rating trade, not a growth trade.
With buybacks MTB could push EPS growth to the 2-6% range near term. Still, low growth stocks can be great investments if purchased well. Throw in a 4.2% dividend yield on the equity and we see 6.25%-10.25% annual IRR’s, even if the stock stays at its current 7.5x P/E multiple.
Below is the company’s forward P/E multiple dating back to 2005.
It is quite rare for MTB to trade below 10x, even during periods of economic turmoil. At 7-8x, investors are seemingly pricing in a 30% decline in EPS estimates vs current analyst figures. If 10-12x is the right multiple, then the market is suggesting that EPS falls to $10-12. That seems a bit overdone.
On the NIM normalization side, if NIMs settle back to 3.75% from 4.06% today (what management generally has been guiding to), then the impact to EPS there would be an incremental fall in EPS of $2.36.
That takes “normalized” earnings to about $15 per share.
Comps and Valuation
Below are the larger regional banks, plus Wells Fargo.
It is quite clear that every bank is incredibly cheap. This group has de-rated from 12.5x earnings to under 7.0x today. This is the cheapest that the sector has ever been (according to Barclay’s bank analyst, who started covering banks in the 1990s on the sell side).
Citi had this to say about MTB recently: “M&T Bank was the most disciplined with its excess cash during the period of low rates and has the strongest capital position among the regionals.”
If $12-18 is the right EPS range, and multiples take 2 years to recover to 10x (not fully giving them credit for 12.5x), then $120 to $180 seems a fair valuation range.
That suggests a large margin of safety at M&T bank.
On the upside, on 2025 figures, M&T could trade back to $180, which still happens to be below its 2022 highs.
Conclusion
MTB is not out of the woods by any stretch. Earnings are likely to retreat, and the prognosis is tough in commercial real estate. Still, the 4.2% dividend yield is likely quite safe at MTB. Their payout ratio on 2022 earnings is a conservative 36%. We do anticipate more noise surrounding the stock with respect to its CRE portfolio but think most of that risk is already priced into the shares.
In 2020, the stock fell only briefly to 8x earnings before bouncing back to 10x. Concerns surrounding their hotel and retail loan book then were extreme.
As we see offices staying vacant and CMBS bonds in steep decline, offices are now the concern du jour. We think a markdown of these is warranted, but with the stock down $65 from its highs last year (and an office portfolio that marked at 50c on the dollar only equating to a $12 hit to book value), we feel reasonably confident that M&T can weather the storm here.
We own both the preferreds and the common.
Thanks for reading! Check out our Marketplace service entitled Cash Flow Compounders: The Best Stocks in the World.We look for high quality, high return on equity, free cash flowing stocks with a proven track record in compounding earnings at higher than market rates. We do deep dives and provide our BEST 2-4 ideas per month. My picks going back to 2011 have produced just under 27% annual returns, putting me within the top 1% of bloggers (TipRanks). Sign up for a free 2 week trial to get our latest ideas!
