$BOH · Bank of Hawaii
Latest call · 2026-04-19
The call
Head-to-head · Claude vs the Street vs reality
| Claude | Street | Actual | |
|---|---|---|---|
| EPS | $1.39 | $1.33 | $1.30 |
| Revenue | $0.20B | $0.19B | $0.19B |
| Direction (1d) | 🟢 UP | ⚪ — | 🟡 FLAT |
| 1d move | +2.0% | — | -0.0% |
| 3d move | +1.0% | — | — |
| EPS error | $0.090 | $0.030 | Street closer |
Thesis
Beat pattern intact — NII tailwind clear. Upside capped by: resistance at $80-81, new CEO first call, and Hawaii flood credit risk against 93% Hawaii loan concentration.
What would flip it
Credit-provisioning spike on flood exposure trumps NII strength.
The market's narrative
NIM expansion + Hawaii tourism recovery drive a clean beat-and-raise story.
Where the Street may be wrong
- 93% Hawaii loan concentration + 2026 Hawaii flooding → credit provision may tick up more than modeled
- New CEO's first print — tone on credit/guidance could surprise either direction
Peer read: Regional bank peers (WTFC same-day, RF last week) showed NIM expansion — supportive setup
Reasoning
- 4 straight quarter-beats — pattern strongly suggests another beat vs $1.33 consensus (last Q: +11% beat)
- NII forecast $150M vs prior-year $127M — 18% growth from rate-cycle tailwind
- Stock at 30-day high $80.09, volume doubled — positioning leans long into print
- Technical resistance: double-top at $80-81 on TV patterns = upside capped near-term
- New CEO wildcard + Hawaii flooding credit risk cap both confidence and 3-day move
Risks to the call
- If credit provisions spike on Hawaii flood exposure, EPS misses despite NII strength
- Guidance cut on NIM from projected rate cuts could fade any beat pop
Lesson from the post-mortem
EPS estimate $1.39 missed the GAAP print of $1.30 — normalized was $1.39 so Claude's number matched the underlying, but the headline (GAAP, which is what the Street print uses for the duel) missed by $0.09 vs Street's $0.03 error. Direction call was UP +2% and actual was flat (-0.04%). The resistance-at-$80-81 call was right and capped the move; we should have sized the UP call smaller (+0.5% to +1%) rather than +2% given the technical ceiling. Lesson: when technicals flag a tight ceiling, don't call +2% direction — call 'FLAT with small UP bias'.