Description

Summary Thesis

The regional bank ETF (KRE) has begun to rally lately, but is still only up 8% in the past year, vs. the S&P up 17%.  With a vaccine, there remains material upside among financials, particularly as the economic cycle returns to growth in 2021. With a stimulus bill lined up, and the Fed keeping rates low, we see the backdrop for a steepening yield curve in place. Throw in diminishing loan loss provisions, and conditions appear ripe for bank fundamentals to improve dramatically.

They are still very cheap too. The S&P at 23x 2021 earnings is stretched, but many banks trade at 9-13x earnings, record high 40-50% discounts to the market. With good cash flow and solid long term loan growth characteristics, many smaller banks appear to be a good trade.

We screened through about 100 regional and community banks, and came up with Financial Institutions (FISI) out of New York, as it boasts a combination of 1) solid ROE’s at 11% over the past 3 years, 2) management who favors organic growth, 3) positive profits in 2008 and 2009, and 4) still trades at only 8.9x 2021 earnings.

Given their loan deposit ratio (82%), and the improving rate/NIM picture, this seems a classic case of the market ascribing a trough multiple on trough earnings.  Before the pandemic, FISI typically traded at a 14.4x earnings multiple.

FISI has never been written up on VIC.

The company saw at least half a dozen insiders buying shares this year, with zero insider sales, and also intends to buy back 5% of the outstanding shares over the next year. 

Their track record is pretty good back to 2004:

In sum, based on what we expect of roughly $3.13 in EPS in 2022, and using a comp multiple of 12x, we calculate upside potential to $37 per share, which with dividends works out to almost 50% upside over the next 12-18 months. Given that clean earnings in Q4 (tax adjusted, excluding one time items), was $0.75, the $3 case for earnings seems quite doable.

Our bull scenario is one whereby loan loss provisions drop to under 2019 figures, as the company has over-reserved for losses, and at 13x the stock trades to $45. 

Our downside case is $20.

Capitalization

The company’s capital ratios are solid. Tier 1 capital is 10.6% as of September 2020, and their leverage ratio is 8.25%. These have remained steady for the past five years.

In late 2008, the Fed force fed all US banks TARP capital in the form of preferreds. FISI accepted $37.5mm of prefs. They repaid that in 2011 with a $40mm equity offering that was 25% dilutive.

Had TARP not been forced on the company, the company could have built up the capital needed in a couple of years. EPS would have shown even better long term growth too. The five-year EPS growth rate (2014-2019) was 8.2%, which is pretty good.

History

The predecessor to Financial Institutions (FISI) is Five Star Bank (formerly Wyoming County Bank), founded in the 1850’s. Wyoming Bank merged with a couple other smaller banks, changed names and went public in 1999 at $15 per share. That was 11x earnings at that time.

Right out of the gate, FISI acquired a few regional players, and with a de-centralized and yet aggressive lending strategy, got into trouble. Loans almost doubled from 1999 to 2002, but poor underwriting led to issues. With a big agricultural loan book at the time, FISI also struggled when milk prices collapsed.

By 2003, FISI was forced to revamp their lending practices and improve their underwriting. Operations became more centralized, lending officers better trained and a Chief Risk Officer was put in place.

Revenue stagnated during that time, for three years, and EPS fell from $2.23 in 2002 to a low of $0.28 in 2005. But an improved underwriting team fixed their loan quality issues, and by 2008 and 2009, led to solid loan performance despite the Great Recession. FISI got smart in time to avoid subprime lending and the no document loans that plagued much of the industry.

As mentioned above, FISI did take TARP money, and paid the price of dilution to redeem them. But since then, the company has run a tight ship from an underwriting perspective.

The original CEO, Peter Humphrey, served in that role from 1983 until retiring in 2012. The “new” CEO, Marty Birmingham, then took over and has stuck to the company’s core rural lending businesses, as well as growing from a rural bank and one with a growing presence in Rochester and Buffalo (their headquarters are located almost exactly between them).

They have done a few very small deals:

SDN was acquired in 2014, and offers insurance brokerage services for customers.

Courier Capital and HPN Capital were small transactions in 2016 and 2018 respectively, and offer private wealth management services.

Management is quite focused on adding more valuable and stable, fee-based revenue.  Recently in fact FISI announced another small deal to buy a Rochester based insurance broker.  Terms were not disclosed.

Business

Today, the company operates 47 branches (down from 53, as some are being consolidated this year). Some key facts:

The company’s loan book is divided into three main categories.

Commercial loans are the biggest category, at $2.0BB as of the end of December. That is 56% of the total, with $1.2BB of that related to secured, mortgage loans.

Consumer Indirect is category number 2, with $840mm of loans outstanding (24% of the total). These are auto loans, but nothing material in the subprime category. The average FICO score of their auto loans (made via partnerships with local car dealers), is 723.

The company has been de-emphasizing their indirect lending since the end of 2018, when the auto lending space got pretty overheated. They also are more focused on direct relationship lending. Knowing their customer that is. Overall, the company has been in this business since 2005, with a very solid track record. Today, management has commented that they like spreads today, and it could be a driver of better loan yields in 2021.

Residential real estate lending is the third biggest lending category at just under 20% of total loans. This is a solid business, with high quality borrowers (769 FICO scores), and growing 6.2% per year over the past five years. The company experienced limited losses in 2008-9, and to date has only $6mm of mortgages on a deferral program, out of a total of $691mm (under 1%).  There is refi risk here, but overall this is a much smaller portion of the balance sheet than it would be at most smaller banks.

Rounding out the company’s balance sheet is $807mm of securities (publicly traded Agency bonds for the most part). With loan yields at 4.25% today compared to securities yielding 2.4%, the company has been migrating capital away from securities and into loans, particularly on the commercial side. This has helped offset net interest margin pressures.

Below is a chart of NIM over the past decade, including Q3.

NIM has remained quite stable since 2015. Even in Q4 2020, NIM was only down 8 bps, but largely driven by lower NIM PPP loans.  In 2020, guidance suggests that it will remain flat. At this point, there is likely only upside to interest margins.  Any whiff of inflation that pushes the 10 year higher, with the Fed keeping the short end low, and banks could rally significantly.  

In 2009, the company’s loan portfolio performed admirably, but at the time FISI owned a portfolio of agency equities and preferreds, and took a hit on that.  Today their securities portfolio only owns Agency backed bonds (with no equities).

Here is a chart of loans ($ in millions) going back 10 years, with an annual growth of 10% per year.

Given the lack of losses in 2020, with 2020 EPS $2.11, underwriting looks fine.

Asset Quality

Banks in general are trading at still large discounts to the market multiples based on loan worries (and lack of growth, only growth stocks performing).

Unemployment is the biggest driver to delinquency levels, and with unemployment levels extremely elevated, defaults seem guaranteed.

Overall, it appears that FISI is pretty conservatively booking reserves for loan losses.

Above is our math on how much the company has set aside to cover future loan losses ($52.4mm).  Compared to non-performing loans, allowances are a very conservative 420%. I consider anything over 120% to 130% to be “good,” so these are conservative on the surface. If we add in loans on a deferral program ($90mm, as it has come down from $99 as of December per management) plus NPLS of $12mm, then we have $102mm of total potentially problematic loans.

First of all, as a percentage of total loans, only 2.9% of loans are either deferred or non-performing, with most of that from their loan deferral program. The good news is that loan deferrals totaled $525mm (16%) as of June 30, but today are only $90mm (2% of loans outstanding).

If we assume 60% default rates and 60% recoveries, then we get loan losses of $37mm.  Against $52 in allowances, the odds look pretty good for reserve releases in 2022.

We honestly have no idea how defaults shake out, but we make the point that assuming employment levels continue to improve, the company is well reserved for future losses in their loan book.

On the Q3 call, management actually stated that had they followed their quantitative models (which are heavily tied to unemployment levels), then they would have released reserves. That is, they would not have booked $4.0mm in provisions for loan losses (which hit the income statement of course). 

Western New York Economy

There is regional risk investing in a bank based in Western New York. But over time, the economy there has continued to grow, housing is quite affordable, and unemployment quite similar to that of the broader US economy.

GDP Growth Since 2002

Unemployment Data

At 7.5% unemployment in December in Western NY, it is a bit higher than the US average.  As of October it matched the US, but generally over time it has followed national trends. 

The economy is pretty diversified, and Cornell projects that the population will remain flat over the next five years in Western New York.

This report suggests that 10 out of the biggest 12 industries in Western NY experienced job growth over the past 5 years.  Many of these are in industries poised to recover.

2021 Guidance

The company gave very specific line by line guidance for 2021. They assumed no steepening in the yield curve (NIM’s constant at 3.13%), loan growth of 5%, and growth in non-interest income in the low single digits.

As for expenses, management also guided to $112mm for the year ($27-29mm per quarter), which is 2% higher than 2020. Overall, FISI expects their efficiency ratio to range between 57-58% for 2021. Banks below 60% in this environment I consider well run. With Q4 Efficiency ratios at 56%, this seems pretty doable, and perhaps beatable with any NIM improvement.

The big question mark for 2021 will be provisions for loan losses. They guided to charge-offs of 35-40 basis points as a percentage of loans outstanding. So, if we assume loans grow 5%, then charge-offs in 2021 probably hit $15mm.

On the Q4 call the CEO said,

"The challenge with -- so no, we didn't provide any specific guidance on provision. We just stuck with Charge-offs. And, the reason is, the same story that we've had this year is, tell me what our unemployment forecasts are going to look like. And tell me what's going to happen with the pandemic. And I can probably do a little bit of a better job, predicting the CECL model outcomes. It's a really tough thing to predict. I, I am very comfortable with where our ratios are right now, from a coverage perspective, from a reserve perspective. But, that doesn't necessarily mean that what happens next year, relative to modeled outcomes and CECL, that doesn't necessarily mean we're going to stay at that reserve level. We could be driven to increase it depending on what happens with economic forecasts. And we could be driven to bring it down depending on what happens to those forecasts. So it's just a really difficult thing to predict. So we've chosen to sort of give you some guidance relative to charge-offs, but haven't really talked much about what's going to happen to the overall provision or allowance."

So, they don’t know. But with vaccines rolling out, and delinquencies actually declining in Q4 (from $13.9mm in Q3 to $12.5mm), the trend is going in the right direction.

Now, loan deferrals remain elevated, at $99mm in Q4, but “lower today” per the company (we assume they are $90mm). Assuming a 60% recovery on 60% of that defaulting, then that implies total loan losses of $37mm. With reserves now built to $52mm, it seems pretty likely that provisions are not going higher.

That said, we assumed losses at the elevated Q4 runrate for all of 2021.

That still gets us $2.86 in EPS for 2021.

For 2022, we assume that provisions fall from $22mm to $12mm. In 2018 and 2019, loan loss provisions were around $8-9mm. So we think 2022 looks conservative.

That gets us $3.13 in EPS for 2022.

Here is our model annually. There hasn’t been a Street EPS update since Q4 figures were reported, so current estimates appear far too low at $2.55 in 2021 EPS.

Our income statement is below:

Community Banks Comps

On average, these comps today trade at 11.1x 2021 earnings, and 11.3x 2022 earnings (using median).

M&T Bank (MTB) is based in Buffalo, but with $79BB in assets, is 20x bigger than FISI. It trades at 12x 2021 earnings, and at a 3.1% dividend yield, seems like a potential buyer of the company at some point.

Historically FISI has traded at a sub 3% dividend yield. Today it is 4.1%, not because of dividend cut concerns, but simply as the market remains wary of banks and value equities.

FISI tends to trade at 14.4x earnings, looking back 5 years pre pandemic.

At 8.9x 2021 and 8.1x 2022, FISI is 30% cheaper than peers today, and well below is normalized trading level.

Here is our valuation on a P/B basis as well as on a P/E basis:

Big picture, we assume:

1. 5% loan growth over the next 3-5 years, plus

2. Share repurchases taking out 2-3% of shares annually, plus

3. Cost cutting of 1% per year.

That implies 6-8% EPS growth long term.  Runrate EPS is $3.00 (from Q4), a reasonable estimate perhaps for the next few months, pre-vaccine. A 13-14x multiple or 1.2x book multiple seems reasoble for 7% earnings growth, double digit ROE bank with takeout potential.

The company has a buyback plan to purchase 5% of the share base over the next twelve months.

Insider buying has also been impressive in 2020, with the CEO and a handful of executives picking up 13,000 shares in the open market back in May (between $18.50 and $19.50).

Risks

Liquidity / Small Cap Name

Regional Economic Dependence

Interest Rate Risk

Acquisition Risk

Loan Losses amidst further lockdowns

Catalyst

Earnings, vaccine, stimulus