Thor Explorations Ltd. THX

Deep-value analysis: THX — Thor Explorations Ltd.

Date: 2026-08-15 · Ticker: THX (TSX Venture, primary) / THX (LSE-AIM) / THXPF (OTC Pink, illiquid) Price: CAD 1.070 (THX.V, close 2026-08-14, Yahoo Finance) · … (THX.L, same date) Rate: CADUSD 0.7209 · GBPUSD 1.3536 (Yahoo, 2026-08-15) Shares outstanding: 666,573,136 (Note 11, interim Q2 2026 statements) Market cap: CAD 713.2 mil. ≈ USD 514.2 mil.

Dual-listing check (market cap cross-check): … × 1.3536 = USD 0.7716/share; CAD 1.070 × 0.7209 = USD 0.7714/share. The two markets differ by 0.03% — the USD 514.2 mil. cap is confirmed independently, not derived from a single quote.

Methodological note — why there is no SEC filing. edgar_10k_downloader.py THX was run today twice — during the initial drafting and again during the final check — and returned each time ERROR: CIK not found for 'THX' — not a US-listed company?, with the output directory Agent-analizor-bursa\SEC-Filings empty for this ticker. Thor is a British Columbia company, listed on TSXV + AIM, reports IFRS in USD and files on SEDAR+ / RNS, not EDGAR. There is no 10-K or 20-F. The primary source for this report is the condensed consolidated interim financial statements for the quarter and half-year ended June 30, 2026 (authorized by the board on August 10, 2026, published August 11) and the associated Q2 2026 MD&A, downloaded today directly from thorexpl.com, plus the audited FY2025 statements (BDO Canada, signed April 8, 2026). The prior deep reports (July 20 and 26, 2026) were based on the operational update from July 15, explicitly without complete financial statements — hence a significant part of the divergences flagged below.

Relationship to the research brief. The brief research-THX-20260815.md (sonnet-5, 14:47) was read in full and treated as testimony, not as truth. Six of its figures proved wrong when checked against the primary filings; the corrections are listed immediately below and repeated in the chapters where they matter.

Corrections to the research brief (each verified against the primary filing)

Claim in the brief	Reality per the filing	Source
FY2025 EPS = US… ("figure with confirmed direct source")	US… basic and diluted	Audited FY2025 statements, p. 8
TTM: revenue US…M, profit US…M	Those are figures in CAD. In USD: revenue 330.6M, profit 205.6M	calculated from FY25 + H1'26 − H1'25
The stock rose ~270% in 12 months	… in 12 months; … YTD; … from the January 2026 peak (…)	Yahoo THX.V, daily series
Cash position ~US…M / adjusted net cash US…M (divergent sources)	Cash 193.128M; "adjusted net cash" 218.648M = cash + 6,367 oz of bullion marked to market (25.520M). The 225.6M figure from the operational update was revised downward	MD&A Q2 Note 3.4, including footnote 2
Douta "completed in 2025," 100% owned	The purchase of the two 30% minority stakes was still unclosed as of June 30, 2026 (ministerial approval pending); Bousankhoba is 70%; the Senegalese state holds 10% free-carry	Note 9a, Q2 2026 statements
Net profit Q2 = US…M (flagged "no individually verifiable URL")	Confirmed exactly: USD 48,736 thousand	Q2 2026 statements, p. 5

The operational figures in the brief (19,153 oz poured, 17,050 oz sold, realized price …/oz, cash cost …/oz, AISC …/oz, 240,769 t at 2.57 g/t, 93.3% recovery) all checked out, to the unit, against Table 2.1 and Tables 3.1–3.3 of the MD&A. The brief was reliable on the operational side and unreliable on the financial side — the typical pattern for third-party aggregated data that mixes the reporting currency (USD) with the quoting currency (CAD).

---

Executive summary

The thesis, in one sentence: Thor Explorations is the cheapest company in Radu's tracker measured by cash flow — the market is paying 1.9 years of current FCFE for the operating business — but it's cheap because it's a depleting asset, not a franchise: the Segilola pit is, in management's own words, "in its final phases," ore grade has fallen from 3.02 to 1.54 g/t in four quarters, and the remaining book value of the depletable asset implies ~109 thousand ounces of reserve, i.e. under a year and a half of production.

The basic arithmetic, from the filings:

Indicator	Value	Derivation
Market cap	USD 514.2 mil.	666.573 mil. shares × … × 0.7209
Cash (hard)	USD 193.128 mil.	balance sheet 06.30.2026
Financial debt	USD 0.158 mil.	current lease only
EV	USD 321.0 mil.	market cap − cash
Net profit TTM	USD 205.6 mil.	FY25 + H1'26 − H1'25
CFO TTM	USD 213.9 mil.	same
FCFE TTM	USD 169.2 mil.	see FCF bridge
P/E TTM	2.50×	gold peers: median ~14×
P/B	1.12×	peers: median ~2.30×
EV/EBITDA TTM	1.31×	peers: median ~6.53×
EV/FCFE	1.90×	—
Adjusted net cash / price	42.5%	… of …

Estimated value — a range, not a point (five models, detailed in the valuation chapter):

#	Model	Value/share	MOS
1	Run-off DCF, gold …/oz, no underground, no Douta	…	…
2	Sum-of-parts NAV, Douta risked at 30%	…	…
3	Base DCF: underground + Douta built, gold …/oz	…	…
4	EV/EBITDA 3.5× on normalized EBITDA	…	…
5	Greenwald EPV with finite life (8-year annuity)	…	…

Median: …. The Monte Carlo simulation over 20,000 scenarios gives a median of … and P(undervalued) = …% — consistent with my models, but be careful what that …% means: the simulation varies the financial assumptions (owner earnings, growth, discount), not the binary event that matters most — is there or isn't there reserve after 2028. The real probability of loss is greater than 0.7%, and model 1 shows how: at gold of …/oz and no underground conversion, the stock is worth 21% less than today's price.

Verdict: MONITOR (55.8%), down from INTERESTING (…%) in the July 26 report. The 13.4 percentage-point drop doesn't come from a methodology change or a price correction — it comes exclusively from the Q2 2026 financial statements, which prior reports didn't have. They turned three criteria from "uncertain" into "proven negative": moat durability (B12: pit in its final phase, book reserve ~109 koz), market share (L28: production declining on guidance, while all peers are growing) and SBC dilution (L30: the RSU plan was reactivated in January 2026, after a year at zero). The full detail, criterion by criterion, is in the last chapter.

How "median MOS …" fits with "MONITOR": it's not a contradiction, it's exactly the Graham–Buffett distinction. The stock is cheap in the Graham sense (a liquidating asset trading below its own value) and not cheap in the Buffett sense (there's no franchise to buy). 42.5% of the price is net cash; the rest is a mine with ~2 years of visible life left and an option on Douta. Whoever buys THX is buying a fat cigar butt, not a compounder — and position sizing must reflect that.

---