Description

Signature Bank has been written up 4 times on VIC, most recently in June 2021.  We recommend reading the latest one for some great background information.  At the time, EPS was run-rating around $12 (Q1 2021), crypto deposits were starting to ramp up (with $13 billion in crypto deposits from $1 billion in 2020), and loan growth was up 25% (organically).  The company did almost $15 in EPS in 2021 and the stock performed quite well. 

In 2021 and 2022, the narrative shifted that SBNY had become a “crypto” bank, bringing with it all the volatility of a high-risk cryptocurrency equity.  Momentum players pushed the stock up to over $350 in early 2022.  Since then, the stock has plummeted to $152, a lower valuation than at the depths of 2020.

This is the forward 12-month P/E chart on Signature Bank.  In a matter of months, SBNY has de-rated from 20x to under 7x forward earnings.

We think the market is overly concerned about a recession as well as the impact of crypto deposit flight risk.  We get it that loan growth can slow and deposits are seeing competition from various funds with higher rates.  But the company has a solid balance sheet and an excellent underwriting track record over time.  The company has $74 billion in total loans, $81 billion in dollar deposits, and $21 billion in crypto deposits.  They have a miniscule $100 million in loans backed by crypto deposits (0.15% of the total), so in the extreme scenario that 50% of their crypto deposits disappear, the impact would solely be on a reduction in their cash and Agency portfolio.  We estimate that impact as roughly $2.50 in EPS (see below).  

EPS was $18.67 on a TTM basis and $5.41 in Q2.  2022 estimates are around $22.

Amazingly, crypto values have fallen 65% year to date, and their crypto deposits are only down 10%.

Rough math, if we take 2022 estimates of $22, cut them by $3, so at $19 without a crypto deposit business, SBNY is still only trading at 8.3x earnings.  We target upside of $200 short term (1 year) and $350-450 longer term, using only a 14x multiple (well below its 16.5x historical multiple).

SBNY has a solid track record of organically growing their loan portfolio by 20% per year from 2010 to 2021.  It historically has traded at a large premium to its banking peers given its track record so we think 14x is conservative once past recession fears/deposit concerns.

As for a recession, we note that since 2010 ROE’s have ranged from 10.1% to 18%.  In a worst case scenario, with ROE’s falling to 10%, EPS falls to $12.  This appears to be the markets implication.  

In our base case, EPS recovers from any recessionary impacts in 2023 and reverts back to trend.  We get ~$26 in EPS in 2024.  At 14x, too low possibly given the high quality nature of this bank, SBNY is a $370 stock, up 140% in 2 years.  We initially purchased shares in 2020, still own them and plan to hold them perhaps for 5+ years.  SBNY has been a solid compounder for investors over its 18 year life as a public company.

Crypto Exposure

SBNY stock took a hit when TerraUSD broke the buck back in May.  Many feared that Signature banked TerraUSD and so might have deposits that would quickly vanish.  It soon became clear that Signature had no exposure to Terra.  But the damage was done.  In June SBNY fell further amidst market carnage, and in September the same happened despite positive news with respect to deposit levels (see their 8K from September 2).  Here are recent disclosures on crypto and dollar deposits.

Signature Bank’s crypto business is called Signet.  They have $21 billion in deposits from miners, funds that want to invest in cryptocurrencies, and exchanges (among others).  SBNY banks 7 of the top 10 crypto exchanges.  Also, about $7 billion of their deposits are related to stable coins.  Based on our conversations with the company, we believe these are TrueUSD and USD coins.  Here is a link to TrueUSD (TUSD): https://www.trueusd.com/.

The latest TUSD price as of October 6th is $1.00 and it remains fully backed by US dollars.  USD coins also trade at $1.00 and are backed by actual US dollars.  TerraUSD failed as it wasn’t actually backed by dollars, but rather would mint or buy back TerraUSD’s based on trading prices.  They held some bitcoin reserves (on a partial basis only), so the reality is that Terra was not a real stablecoin.  SBNY mentioned that they would never back something with such poor credit metrics.

This above assumes $10.5 billion in lost deposits from either crypto or in the event that their $41 billion in non-interest bearing deposits migrate away from the company to higher yielding instruments.  Deposit competition is fierce right now.  

Given the strength of their balance sheet, we aren’t too worried even if the company were forced to sell up to $10 billion in Agency paper in order to keep the balance sheet at required levels of cash. The $10 billion figure is one the company suggests is a minimum, but they have acknowledged it could drop below that a bit.

Even if this did happen, management suggested that they could easily raise debt to fund any cash needs.  So, we think the above is pretty conservative.  Their bonds today yield 5.15%.

Here is the company’s Q2 slide on deposit trends.

Recession

 Below are non-performing assets as a percentage of total loans. 

Here are long term Returns on Average Equity.

Returns fell in 2017 on losses on their book of NYC Taxi Medallion loans.  Returns were recently boosted by pandemic low defaults/delinquencies, reserve releases in 2021 and by low Fed Funds/deposit costs.  We’d call 10% the low end and 14% a normalized ROE.  We think ROE’s will remain higher than long term averages given the rate picture and the companies steady improvement in their efficiency ratio.

Efficiency ratios are incredibly low at 30%.  This is an extremely well run bank.

To put some estimates around their loan loss potential in a recession, we model loan losses quadrupling from current levels (non-accrual loans as a percentage of total).  On 2 occasions since 2005, non-performing loans hit 1.0% of total loans (in 2009 and again in 2017 after their Taxi medallion losses).  Their 10 year average is around 0.3%. 

If a recession drives losses back up to 1.0%, then 2023 EPS could be impacted by about $5.  In 2020, EPS only fell 6.1% year over year and in 2009 EPS fell 18.6% (from $1.67 to $1.36).  So, EPS dropping from our $21 estimate to $17 (down 19%) appears quite conservative.  We don’t view a Fed induced recession as one that hits credit as hard as the 2004-2008 housing/lending bubble.  For comparison’s sake, Street estimates are $21.51 and $23.74 in 2022 and 2023.

Here is our summary income statement model.

We admittedly are guessing here, but assume NIM’s fall to 2.0% and loan losses spike, with provisions for loan losses roughly doubling from 2020 levels.  With a huge pile of non-interest bearing loans and 52% of their loans are floating rate, the company is asset sensitive and should benefit from higher interest rates.  NIM increased from 1.99% in Q1 to 2.23% in Q2.  That could offset a lot of loss provisioning in a recession.

The company per the above suggests that NIM will improve dramatically with higher rates. We are frankly skeptical given deposit competition, but think that our 2.0% figure is extremely conservative.

Long term loan growth has been impressive, and the company continues to add banking teams in new cities like Chicago and Dallas as well as continuing to penetrate in California.  Loan growth has lately been boosted by a category called Fund Banking, which are loans to private equity funds.  Instead of calling capital, PE firms are now borrowing against LP commitments and target purchases in order to boost their own ROE’s.  This could slow as fundraising in PE land slows, but long term we aren’t calling an end to that growth angle.  The concern is that at high enough rates, PE firms will utilize these facilities less than otherwise.

From a valuation perspective, with rates rising there are worries that their deposits could decline on top of a recession.  Our severe recession scenario is a collapse in ROE’s to 10% (despite the likelihood that higher rates improves net interest margins).

Our model above assumes a recession in 2023 already in our base case.  We point out that long term Signature Banks has been an impressive compounder.  EPS has grown at a 20% CAGR over a 15 year period through 2 credit cycles.  Their growth has been entirely organic and they have yet to have a negative earnings year since its founding in 1999. 

In 5 years, assuming loan growth of 10% annually (compare that to 18% prior loan growth from 2010 to 2021), plus flat NIMs, a 1.5% dividend yield, plus 1% annual margin improvement and we end up at 80% higher cumulative earnings per share.  If you base that from 2021 and 2023, then $27-31 in EPS is quite possible.  At 15x, that gets us $400 to $460.  Annual IRR’s would be 21-25%.  Our near term downside case is $120.

Catalyst

Earnings, Fed pivot will happen at some point, underwriting