Dirt Cheap Banks · Mar 18, 2026 at 08:08 PM
Done some work on this one. A few thoughts:
On the bank itself, for context for others reading: this is Citizens Financial Corp. (CIWV), a ~$686M asset community bank operating out of Elkins, WV under CEO Nathaniel Bonnell. It's been recognized four consecutive years as a Top Performing Publicly Traded Community Bank Under $2 Billion in Assets by American Banker, and is ranked #1 as the consumer/mortgage lender in West Virginia in its asset class. Very thinly traded OTC name -- probably why the valuation looks as cheap as it does.
The growth story is real. When Bonnell took over, assets were around $340M with five locations. They're now at $686M with six financial centers, and recently announced a merger with Miners and Merchants Bank to expand the footprint further. That's roughly a doubling of the balance sheet, meaningful execution for a bank this size.
On stress testing the asset side:
The consumer-heavy book (auto, RV, boat, personal loans, mortgages) gets analyzed differently than CRE-heavy community banks. The key things I'd pull from the annual filings:
NCO history through 2008-2010 and 2020. Those are your two real stress tests. If the book held below ~1% NCOs through either cycle, that's your underwriting evidence. Current credit quality looks pristine, nonperforming assets of just $2.7 million on a $544M loan book, with the allowance sitting at $5.4 million. That's under 0.5% NPA/assets, which is excellent, but cycles haven't been stress-tested under Bonnell's growth-driven origination strategy yet. That's the gap to fill.
The collateral mix matters a lot here. RV/boat/auto loans are secured consumer, materially better recovery rates than unsecured personal or credit card. Loss severity on a repossessed bass boat is very different from a charged-off personal loan. Most analysts lump "consumer" together and that's a mistake.
The NIM expanded dramatically to 4.05% for the nine months ended September 2025, up from 3.28% in the prior year period. Running a basic stress scenario: if you assume 1% normalized NCOs (which would be elevated vs. history for secured consumer), you're still working with 3%+ spread before opex.
On whether competitors can come in and compress the asset-side ROE:
This is the right question and the honest answer is: somewhat, but less than you'd expect. Here's why the moat is stickier than it looks:
The yield advantage isn't purely a pricing decision, it's an origination distribution advantage. Citizens serves customers across Randolph, Tucker, Upshur, Pocahontas, and Wetzel Counties. These are rural Appalachian markets where the #1 barrier to lending isn't price, it's access and trust. A regional bank coming in to compete on auto loan pricing has to first build dealer relationships, brand recognition, and local underwriting knowledge from scratch. CIWV already has that infrastructure and the #1 market position. The loan flow comes to them.
Consumer borrowers in these markets are also stickier on rate than corporate borrowers. A 30bps rate difference on a $25K RV loan is ~$6/month. That's not going to move many borrowers away from a bank where their family has banked for decades.
The real secular risk isn't regional bank competition, it's fintech (LightStream, Marcus, etc.) competing nationally on unsecured and auto with zero branch overhead. Worth monitoring what % of the book is truly relationship-originated vs. potentially disintermediated over a 10-year horizon.
Bank-specific things I'd dig into:
Loan vintage analysis. Aggressive growth since 2017 means a lot of paper hasn't seasoned through a full credit cycle under the current origination strategy. Pay attention to the 2018-2022 vintage NCOs over the next few years.
The Miners and Merchants integration. Acquired books always carry unknowns -- you're buying someone else's underwriting decisions. Watch the first 2 years of post-close credit metrics closely.
Shareholders' equity grew at a 20%+ annualized pace through Q3 2025. Book value compounding at that rate while the stock trades at ~1x TBV means you're getting ROE for free if earnings hold.
On the valuation, the liquidity discount is doing a lot of work here. This isn't mispriced because the market doesn't understand it; it's mispriced because 99% of capital can't own it. That's the opportunity, and also the constraint on getting out if the thesis breaks.
