Setup

Washington Trust Bancorp (WASH) filed Q1 2026 earnings Apr 20 after close. EPS missed consensus by 13%. That's not the story. The story is one line in the CRE detail table: a $34M "Office Lab Space" loan at 103% LTV went nonaccrual, driving commercial NPLs from $0 at Dec 31, 2025 to $40.4M ninety days later. Three days earlier, Independent Bank Corp (INDB, Rockland Trust) printed criticized commercial +$102.7M QoQ, with a $115.6M medical office subcategory. Same geography. Same quarter. Same asset class.

Two prints don't make a trend. They might identify the next one.

What the filing says

WASH Q1 2026:

• EPS $0.66 vs $0.76 consensus (-13%). Driven almost entirely by a $3.4M provision swing ($0.6M net recovery Q4 → $4.0M Q1).
• NIM 2.63%, +7bps QoQ. Balance-sheet repositioning thesis intact.
• Commercial nonaccruals $0 → $40.4M QoQ. Two CRE office loans.
• The problem credit: $34M lab space loan, 103% LTV — asset is underwater.
• Office CRE book: $231M / 42 loans. Class B/C medical office $89M. Rest of book (Class A/B/C at 54-59% LTV) looks clean.
• ACL/loans 0.73% → 0.82%. Net charge-offs $10K.

Dec 31, 2025 had commercial nonaccruals at $0 and both 2025 problem credits resolved. The lab space loan surfaced between quarters. Management either didn't know, or knew and didn't disclose the watchlist. Apr 21 call will disambiguate.

What the market thinks

The relevant market is not WASH — WASH is repricing today on the headline. The relevant market is Eastern Bankshares (EBC) at $20.97, RSI 83.4, +47.9% 1Y, reporting Apr 23-24. EBC is the largest Boston-area community bank post Cambridge Bancorp merger, with the most direct Kendall Square lab CRE exposure in the cohort. 20+ analysts. No sell-side note has connected WASH + INDB + REIT-side lab commentary.

Market-implied P(EBC clean print): derived from three signals — EBC's +5pp YoY excess vs KRE (outperformer premium), ATM straddle implying ±7.8% 23-day move, OTM put skew at +125% vs ATM (informed flow paying for tail but narrow — 60 puts bought today at $15 strike). Composite: 75-85%.

Our P(EBC clean print): 2/2 NE community banks confirming pattern + REIT-side (ARE Q4 guide of $24M/yr tenant wind-downs; BXP Q4 "wet lab has not recovered"). Offsetting: HarborOne integration gives management cover. 25-45%.

Gap: 30-55 percentage points. Wide. We can't narrow it below 25pp, and that width is itself the finding — the market has not synthesized the cross-ticker pattern.

Why the gap exists

Two mechanisms:

Information is public, synthesis is not. INDB filed Apr 17, WASH Apr 20, ARE Q4 call Jan 27, BXP Q4 call Jan 28. Sell-side covers banks by geography OR REITs by sector — nobody reads banks through REIT tenant-data. The signal requires assembling four sources across two sectors and three weeks.

Cambridge Bancorp M&A noise masks it. EBC's HarborOne integration added $495M to criticized loans in Q4 2025 on acquisition accounting. Any Q1 migration can be framed as "integration cleanup." Management has discretion to aggregate. Our own earlier prediction (65%) says HarborOne cleanup drives C+C down — competing hypothesis that resolves on the same print as lab-space deterioration. Both can be partially true; the headline goes to whichever side management emphasizes.

Risks

1. HarborOne cleanup dominates the EBC headline. Most likely failure mode. Stock rips on clean narrative; lab exposure buried. Mitigant: BRKL and BHLB report 2-3 weeks later with cleaner structural exposure, no M&A noise.
2. Lab space is WASH-idiosyncratic. A single $34M credit, not systemic. Counter: INDB's $115.6M medical office plus ARE/BXP REIT confirmation don't depend on WASH. Four data points from two sectors.
3. Fed cuts 50bps, regional-bank sector squeeze. Sector beta overwhelms idio. KRE at RSI 85 already. Hedgeable via KRE pair.

Catalysts

• Apr 21, 8:30 AM ET (today): WASH Q1 call. Management language on lab-loan isolation is the primary disambiguator.
• Apr 23-24: EBC Q1 print. Binary event.
• Early-to-mid May: BRKL, BHLB Q1 prints. Cycle confirmation window.
• Jul 2026: WASH Q2. Tests prediction (60%: additional CRE office nonaccrual OR NPLs >$35M).

What would change our mind

• EBC 10-K shows CRE office <5% of loans → thesis inapplicable to EBC specifically, pivot to BRKL/BHLB.
• WASH Apr 21 call: lab loan is specific, watchlist clean for remaining office → reduce prediction to ≈35%.
• BRKL AND BHLB both print clean in May → cycle thesis dies.
• EBC prints similar-magnitude credit migration to WASH/INDB → gap collapses, market priced it on Apr 24.

Evidence