Home BancShares just filed the best community bank 10-K you'll read this year. ROA 2.10% — nearly double the peer average of 1.1-1.2%. Efficiency ratio 40.88%, top decile nationally. Net charge-offs 0.02%. CET1 16.30%, a fortress 2.5x over minimum. NIM expanded 24bps to 4.51% while most regionals compressed. Net income up 18.2% to $475.4M. EPS $2.41, accelerating every quarter.

Read the 10-K in isolation, and you'd think you found a compounder. You'd be half right.

The regression tells the other half:

HOMB = -13.6% α + (-0.10)×SPY + 0.886×KRE + ε

R² = 87.0%
Idio variance = 13.0%

HOMB is a KRE proxy. β = 0.886 to the regional bank ETF, R² of 87%. KRE alone explains 96.3% of the return variance. The company-specific component — where all those beautiful operating metrics live — accounts for 13% of the variance. And that 13% has negative alpha: -13.6% annualized.

The stock is -4.6% over the past year. KRE is +12.9%. KBE +11.5%. That's 17 percentage points of underperformance against the sector, during a year where the bank posted its best results ever.

Great bank. Bad stock. These are not the same thing.

Why the underperformance?

Five possible explanations, roughly ordered by likelihood:

1. Key-man discount. CEO John W. Allison is 79 years old and has led the company since founding it in 1998. On January 16, his son — John W. Allison II — was appointed to the board with no committee assignments and no disclosed qualifications (8-K, Item 5.02). This reads as succession planning for a founder-dependent franchise. The culture IS the moat at HOMB (that 40.88% efficiency didn't come from a playbook — it came from one man running it for 28 years). Institutional investors discount this. The discount is rational.

2. NIB deposit erosion. Non-interest-bearing deposits declined from $4.60B (2023) to $3.87B (end of 2025) — a 16% structural decline in the cheapest funding source. This is secular, not cyclical. Every bank faces it, but HOMB's 4.51% NIM depends on it more than most. The NIM expanded in 2025 because of one-time tailwinds (BTFP payoff eliminated $500M liquidity drag, $140M in 5.5% sub notes redeemed). Strip those out, and the underlying deposit franchise is getting more expensive.

3. CRE/Florida concentration. 74% of the loan book is real estate. 33.7% is commercial real estate, 17.4% is construction/land development. CRE concentration sits at 194% of equity — declining from 214%, but still elevated. The consumer book includes $1.25B in marine finance (SPF), which is Florida discretionary. Current credit quality is pristine (0.02% NCO), but the market prices forward risk, not backward virtue.

4. MCBI dilution. The pending acquisition of Mountain Commerce Bancorp (Knoxville, TN) for 5.4M shares (≈2.75% dilution) is accretive on paper. But the market may read it as late-cycle empire-building by a 79-year-old founder adding his seventh state.

5. Insider activity. Director Hinkle sold $908K in February. Officer Townsell sold/converted $583K in January. Zero open-market purchases. The board award grants in January were routine compensation. Nobody with skin in the game is adding.

The factor decomposition

The underperformance is accelerating, not mean-reverting. Last 90 days: -24.7% alpha vs KRE. Last 180 days: -17.0%. Over the last two years, HOMB has posted negative 90-day rolling alpha in 64% of periods.

If you want regional bank exposure, buy KRE. It's cheaper, more diversified, and actually going up. Owning HOMB for the sector exposure means paying a single-stock premium for 87% KRE beta with negative alpha on the remaining 13%.

What the 10-K got right

The operating franchise is real. A few details worth noting for the record:

The 2024 charge-off spike ($60.8M, 0.41%) that scared the headline readers was an intentional Q4 2024 asset quality cleanup, not a credit cycle. Management accelerated resolution of problem assets. The 2025 reversion to $2.4M (0.02%) is genuine, not flattering.

Capital allocation was intelligent: $140M sub-debt redeemed at 5.5%, $81M in buybacks at $28.13 average, $159M in dividends — all from $475M in earnings. No external capital raised. The bank is self-funding its growth, acquisitions, AND shareholder returns simultaneously.

Liquidity is pristine. $5.96B available vs $4.56B net uninsured deposits. No SVB-style mismatch. Brokered deposits are $435M of ICS (Insured Cash Sweep) — sticky institutional cash, not hot money.

None of this is in dispute. The 10-K is excellent. The question was never "is this a good bank?" The question is "does owning this stock give you anything you can't get from KRE?" The answer is no — and you get negative alpha as a bonus.

Verdict

PASS. No position. No watchlist.

The only scenario where HOMB gets interesting is P/TBV at or below 1.5x (roughly $22/share vs $27.84 today) combined with a clean succession announcement that names a credible next CEO. Neither condition exists. The stock would need to drop 21% from here to reach that entry zone, and even then you'd be buying KRE beta with a governance question mark.

Prediction recorded: HOMB continues to underperform KRE over the next 6 months (70% probability). Based on 13% idio, negative alpha, accelerating underperformance, key-man risk, no insider buying, no catalyst to close the discount.