ESEA — Euroseas Ltd.

Euroseas Ltd. (NASDAQ: ESEA) — deep-value analysis

Date: 2026-09-04 · Reference price: … (close 09/02/2026, yfinance) · Market cap: …M · Shares: 7,055,381 (06/30/2026, 6-K balance sheet) Regime: REFRESH over the 07/26/2026 thesis · Trigger: 2 new filings — 6-K 08/14/2026 (Q2/H1 2026 results) and 6-K 08/25/2026 (M/V Jonathan P charter extension)

What is carried over and what is re-derived

Carried over from the baseline analysis (07/26/2026), without re-verification: the business description and the time-charter model, the outsourcing structure to Eurobulk, the industry framework (ton-mile demand, the Red Sea diversion), and customer concentration — the latter was specifically re-verified in the FY2025 20-F (line 1189) and is exactly correct: top-5 customers = 87% of 2025 revenue (OOCL 32%, Maersk 23%, ASYAD 15%, CMA 10%, ZIM 7%), 76% in 2024, 82% in 2023.

Re-derived, because new filings or identified errors moved them:

1. All H1/Q2 2026 figures — didn’t exist at the baseline thesis. Balance sheet line by line 12/31/2025 → 06/30/2026, income statement, cash flow statement.
2. Owner earnings (oe) — recalculated from scratch. The baseline figure of …M was entered into mc_dcf.py as …, i.e. in thousands, when the script requires millions. Result: a median intrinsic value of …/share and MOS of …. A 1,000x unit error, not a reasoning error — but it makes the baseline simulation unusable. My new figure: …M.
3. The growth rate (g1) — from 0.17 to 0.06. The reason is methodological, detailed in the valuation chapter: 17% was the fleet-capacity CAGR, applied to owner earnings from which growth capex had NOT been deducted. That growth costs …M; putting it in the model for free is double counting.
4. Net debt (nd) — from … (the same unit error) to 28 million, including the new minority interest and debt securities as liquidity.
5. The newbuilding program — from “10 ships, ~…M” to 12 ships, ~…M contracted, of which ~…M still unpaid.
6. Related-party fees — from “~…M/year” to …M in 2025, a figure reconstructed from notes 8 and 12 of the 20-F.
7. Two new structures, absent at the baseline thesis: the partnership with NRP Project Finance AS (49% minority interest) and a …M securities portfolio.

The baseline thesis’s falsifiers — all six — hold at the mechanical triage from 09/03/2026 (debt/equity 0.43 vs threshold 0.8; ROE 27% vs threshold 10%; P/E 3.94 vs threshold 12; FCF yield …% vs threshold 5%; P/B 1.05 vs threshold 1.6; operating margin 64… vs threshold 35%). None has broken. What changed isn’t the thesis, but the size of the implicit bet: capital commitment grew 13% and is now 90% of market cap.